freedombone/beaglebone.txt

#+TITLE: FreedomBone
#+AUTHOR: Bob Mottram
#+EMAIL: bob@robotics.uk.to
#+KEYWORDS: freedombox, debian, beaglebone, friendica, email, web server, home server, internet, censorship, surveillance, social network, irc, jabber
#+DESCRIPTION: Turn the Beaglebone Black into a personal communications server
#+OPTIONS: ^:nil
#+STYLE: <link rel="stylesheet" type="text/css" href="index.css" />

#+BEGIN_CENTER
*How to turn the Beaglebone Black into a FreedomBox-like personal communications server*
#+END_CENTER

[[./images/freedombone_small.jpg]]

#+BEGIN_CENTER
Copyright (C)  2014  Bob Mottram

Permission is granted to copy, distribute and/or modify this document under the terms of the [[https://gnu.org/licenses/fdl.html][GNU Free Documentation License]], Version 1.3 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts.

Source for this web site in [[https://en.wikipedia.org/wiki/Org-mode][Emacs org-mode]] format is available [[/beaglebone.txt][here]]. Comments or patches may be submitted via [[https://github.com/bashrc/freedombone][Github]].
#+END_CENTER

* Introduction

#+BEGIN_VERSE
/If you look at it from an engineering perspective, an iterative perspective, it’s clear that you have to try something rather than do nothing./

-- Edward J. Snowden
#+END_VERSE

** What is FreedomBone?
Today many of us rely upon "free" services in the cloud, such as Gmail, Facebook, Google+ and so on.  It might appear that these services are indispensible infrastructure of the modern internet, but actually they're not strictly needed and the amount of value which they deliver to the average internet user is very marginal.  It is possible to be a citizen of the internet and yet not use those things - to disintermediate the most well known companies and cut out their prurient or merely cringeworthy business models.

FreedomBone is a personal home communications server based upon the BeagleBone Black hardware.  It's small and cheap and will allow you to use email, have your own web site and do social networking in a federated way without needing to rely upon any intermediary companies other than your ISP.
** Do I need any prior knowledge?
In these instructions only a minimal level of familiarity with Linux is assumed.  It's assumed that you know the basics of the /nano/ and /emacs/ editors, but it would be simple to also use other editors if you prefer.
** Why should I do this?
You should consider doing this if you are a freedom-oriented sort of person and you want to maintain sovereignty over your information.  Laws in many places in the world consider you to have relinquished any property rights over data which you put onto a server not owned by youself (i.e. owned by a third party, such as Google or Facebook).

If you don't like the idea of having all your communications intercepted and investigated by the Surveillance State then you should consider running a FreedomBone.  If your profession involves maintaining confidentiality as an essential feature, such as legal or medical services, counselling, teaching or any sort of activism then you should consider running a FreedomBone.  Especially if your activities include [[https://firstlook.org/theintercept/article/2014/03/20/inside-nsa-secret-efforts-hunt-hack-system-administrators/][systems administration]] or [[http://www.spiegel.de/international/germany/gchq-and-nsa-targeted-private-german-companies-a-961444.html][software engineering for any communications-related systems]] then it is highly likely that you have already been targeted and "tasked" by the surveillance apparatus.

As Eben Moglen noted in his now famous [[https://www.youtube.com/watch?v=QOEMv0S8AcA]["Freedom in The Cloud"]] talk the simple fact of you keeping your own internet logs (found in the /var/log directory) puts a certain amount of power in your hands and takes it away from parties who would otherwise sell that information without your knowledge or permission to advertisers or other shady outfits who may not have your best interests at heart.
** After it's installed will it need a lot of maintenance?
So long as the hardware is ok the amount of maintenance needed should be very small.  Unlike on Windows based systems you don't need to defragment drives or mess about with anti-virus programs.  I ran a similar Sheevaplug system between 2010 and 2013 with only occasional software updates or reboots, and uptime was probably 99% or better.
** Is it secure?
Nothing is totally secure or infallible.  You could have the most secure technology and yet still use easy to guess passwords.  In general any software described as "uncrackable" or "guaranteed secure" is likely to be bogus and should be treated with suspicion.  No matter what the hype may claim, all software has bugs so it's really a question of whether your communications are more secure or less secure.  Using something like Freedombone will be likely to increase your degree of communications security to a level which is above average.

This system will not defend you from an attacker who is actively trying to block or corrupt your communications, but I assume that doesn't apply in the majority of cases.  Another thing to be aware of is that running a FreedomBone could make you more vulnerable to traffic analysis, since the server is associated with your home address and isn't a giant aggregation of users somewhere in the cloud.  You need to weigh this alongside the additional legal protection which owning the server and having it in your own home gives you.

FreedomBone should be far more secure than using popular cloud-based services which have spying built into them as a core feature (although not one which is typically advertised), but it is not necessarily any kind of impenetrable information fortress.

This project is not only about security.  It's also about having independence and at least in the realm of information being able to have more control over your own life, without having gatekeepers, censors or companies in the middle.  That's the way that the internet was designed to be in the first place.
** Will running a server all the time affect my electricity bill?
Hardly at all.  The BeagleBone Black consumes very little power - less than 5W.  It would even be potentially possible to run it from a solar panel.
** Can I use a Raspberry Pi or Cubieboard instead?
These instructions are not highly specific to the Beaglebone Black and so will likely also work on other single board computers (SBCs) such as the [[https://en.wikipedia.org/wiki/Raspberry_pi][Raspberry Pi]] or [[https://en.wikipedia.org/wiki/Cubieboard][Cubieboard]].  The original Raspberry Pi only had 256MB of RAM and so the performance of some services may be more limited.  The Beaglebone Black was chosen mainly because of its low cost, relatively good CPU performance for the price (by the standards of 2013) and also low electricity consumption.  The Cubieboard is also another good alternative, with the A20 version having similar specifications but twice as much RAM as the BeagleBone Black.
** Why should I trust the packages or source code downloaded from this site?
If you're particularly security conscious then you shouldn't.  Binary or source packages have only been included here for convenience and to avoid confusion.  "/Go and find a Debian installation for the BeagleBone Black somewhere on the web/" is too vague an instruction for my liking, and I've attempted to keep things as concise and unambiguous as possible - particularly with an average or new Linux user in mind.

However, for maximum security for those software systems which are not already packaged within the Debian repositories then seek out the original sources and verify the hashes independently.

It's worth adopting an attitude of "/trust but verify/".  Don't let fear of mass surveillance and [[https://www.techdirt.com/articles/20140207/08354426130/gchq-has-entire-program-dirty-tricks-including-honeypots-using-journalists-deleting-online-accounts.shtml]["dirty tricks"]] paralyse you into trusting nothing and consequently doing nothing.  Doing nothing means that the surveillance apparatus has succeeded in keeping you under observation at all times.
* Inventory

#+BEGIN_VERSE
/You can’t help someone just by making a wish to do so, you have to take action./

-- Dalai Lama
#+END_VERSE

These instructions assume that you have the following ingredients.

** A BeagleBone Black (BBB)

It should come with a suitable USB cable for the initial setup.  To make things look nicer you may also want to get a case for it.

** An internet connection

It is assumed that the most common situation is via a router installed at home. The router should have ethernet sockets on it and a web interface which allows you to forward ports (sometimes under the "firewall" settings), so that you can forward ssh and web traffic to the BBB.

** microSD card

To use as the main storage for the BBB.  16 or 32GB is fine, and can be obtained quite cheaply.  Try to use Sandisk (class 10 or better) where possible and avoid cheaper cards which often have poor performance.

You may also need an SD card adaptor or USB card reader in order to flash the operating image to the microSD card.  For instance, many laptops have an SD card slot but not a microSD slot.

** 5V/2A power supply

With a plug suitable for powering the BBB.  If you have some device with a USB socket nearby you may also be able to just use that for electrical power.  However, powering from the USB cable alone might result in crashes when the system is under load, depending upon how many milliamps can be supplied by the USB hub/socket.  If the system crashes due to running out of power then you will see that the LEDs on the BBB are continuously on, rather than flashing.  One way to test whether the board has enough power is to try compiling a Linux kernel on it, but any CPU and disk intensive program will also suffice as a test.

[[http://beagleboard.org/Support/FAQ][beagleboard.org]] gives the following advice on power supplies:

#+BEGIN_VERSE
/Power over USB is sufficient as long as the software and system running perform some management to keep it under the USB current limit threshold. For simplicity and maximum capability, powering over the 5V barrel connector is typically recommended./

/The power adapter is required to provide 5V over a 5.5mm outer diameter and 2.1mm inner diameter barrel connector (a barrel connector length of 9.5mm is more than sufficient). The recommended supply current is at least 1.2A (or 6W), but at least 2A (or 10W) is recommended if you are going to connect up anything over the USB./
#+END_VERSE

The plug should be /centre positive/, meaning that the centre/tip is positive and the outer part is negative.

** An ethernet patch cable

Just an ordinary cat5 or cat6 cable that you can get from most electrical/computer stores.

* Installing Debian onto the microSD card

The Debian Linux OS will be installed onto a small flash drive.  It's a good idea to do this rather than using the internal flash, because it will allow you to easily create backups of the entire system if necessary using the dd command.

Download the image.

#+BEGIN_SRC: bash
cd ~/
wget http://freedombone.uk.to/debian-7.2-console-armhf-2013-11-15.tar.xz
#+END_SRC

Verify it.

#+BEGIN_SRC: bash
sha256sum debian-7.2-console-armhf-2013-11-15.tar.xz
262ea96d6bff530ad545e001eb2aa50b26a999c02f0c0e2e5f8536edf21c973a debian-7.2-console-armhf-2013-11-15.tar.xz
#+END_SRC

Uncompress it.

#+BEGIN_SRC: bash
tar xJf debian-7.2-console-armhf-2013-11-15.tar.xz
cd debian-7.2-console-armhf-2013-11-15
#+END_SRC

Create the disk image, where sdX is the name of the flash drive (probably it will be sdb or sdc).  An easy way to find out the device name of the flash drive is to enter the command:

#+BEGIN_SRC: bash
ls /dev/sd*
#+END_SRC

then plug in the flash drive and type the same command again.  You'll be able to see the difference.  Once you know the device name then you can proceed to install the image onto the flash drive.

#+BEGIN_SRC: bash
sudo apt-get install u-boot-tools dosfstools git-core kpartx wget parted
sudo ./setup_sdcard.sh --mmc /dev/sdX --uboot bone --swap-file 1024
#+END_SRC

Once completed then safely remove the microSD card via your file manager (usually right click and "safely remove" or "eject").

* Setup

#+BEGIN_VERSE
/Build the tools for a future you would want to live in/

-- Kurt Opsahl
#+END_VERSE

** Things to be aware of
*** A note on ssh
When using ssh to log into the BBB if you get warnings of the type "/the ECDSA host key for domain differs from the key for the IP address/" then run the command:

#+BEGIN_SRC: bash
ssh-keygen -R <IP address>
#+END_SRC
*** Passwords
It's highly recommended that you use a password manager, such as KeepassX, and make all your passwords long random strings.  It's also a good idea to use different passwords for different pieces of software, instead of one or two passwords for the whole system.  That compartmentalises the security such that even if an attacker gains access to one system they can't necessarily get access to others.
*** HTTPS
Throughout these instructions self signed SSL certificates are used to implement access to web pages via HTTPS.  The whole HTTPS security model upon which much of the internet currently rests seems broken in that it usually depends upon "trusted certificate authorities" who are not really trusted, except perhaps by the maintainers of certain web browser software.  So all that HTTPS really guarantees is that you have an encrypted connection, but an encrypted connection /to who/ can be subject to doubt.  As was seen in 2013 with the [[https://www.schneier.com/essay-455.html][information coming from Edward Snowden]], and also the [[http://en.wikipedia.org/wiki/Lavabit][Lavabit email service]], it's possible for companies/organisations to be compromised or bribed and SSL private keys for all users can be demanded using gagging orders or secret laws without any individual user ever being able to know that their communications is no longer secure..

Not knowing who you're really connecting to is especially true for self-signed certificates, so it is in principle possible that when logging into a site with a username and password a system such as [[http://arstechnica.com/tech-policy/2013/11/uk-spies-continue-quantum-insert-attack-via-linkedin-slashdot-pages/][Quantum Insert]], or a compromised [[http://en.wikipedia.org/wiki/Domain_Name_System][DNS service]], could be used to direct the user to a fake copy of the login screen for the purposes of obtaining their login details.  While this doesn't seem to be a major problem at the time of writing it's something to keep in mind.  So if you can't log in or if you log in and what you see doesn't look like your site then it's possible that such a compromise could have taken place.  Using a password manager with different login details for each site is one way to ensure that if one system is compromised then the attacker can't necessarily get access to all your other stuff.
** Initial

Eject the microSD card from your computer and plug it into the BBB, then connect the USB cable between the two.  You may need to wait for a couple of minutes for the BBB to boot from the card, then you can then open a terminal and login via ssh.

#+BEGIN_SRC: bash
ssh debian@192.168.7.2
#+END_SRC

The default password is /temppwd/

Then log in as root:

#+BEGIN_SRC: bash
su
#+END_SRC

The default password is /root/

The first thing to do is to change the passwords from their defaults.

#+BEGIN_SRC: bash
passwd
#+END_SRC

Then you will need to change the network interfaces.  The main task here is to comment out the stuff related to usb0.  That will enable you to plug the BBB into the back of a router and for it to be detectable on the network.

#+BEGIN_SRC: bash
nano /etc/network/interfaces
#+END_SRC

The resulting interfaces file should look like this:

#+BEGIN_SRC: bash
# This file describes the network interfaces available on your system
# and how to activate them. For more information, see interfaces(5).

# The loopback network interface
auto lo
iface lo inet loopback

# The primary network interface
allow-hotplug eth0
iface eth0 inet static
    address 192.168.1.60
    netmask 255.255.255.0
    gateway 192.168.1.254
    dns-nameservers 213.73.91.35 85.214.20.141
# Example to keep MAC address between reboots
#hwaddress ether DA:AD:CE:EF:CA:FE

# WiFi Example
#auto wlan0
#iface wlan0 inet dhcp
#    wpa-ssid "essid"
#    wpa-psk  "password"

# Ethernet/RNDIS gadget (g_ether)
# ... or on host side, usbnet and random hwaddr
# Note on some boards, usb0 is automaticly setup with an init script
# in that case, to completely disable remove file [run_boot-scripts] from the boot partition
#iface usb0 inet static
#    address 192.168.7.2
#    netmask 255.255.255.0
#    network 192.168.7.0
#    gateway 192.168.7.1
#+END_SRC

CTRL-o followed by ENTER to save, then CTRL-x to exit.

In the above example "address 192.168.1.60" is a static IP address for the BBB, which will allow incoming network traffic to be directed from the router in a reliable manner.  It should be outside of the DHCP range set up on the router.

"gateway 192.168.1.254" should be the IP address of the router.

Note that setting the DNS servers with dns-nameservers is important because some home routers do not allow you to change the DNS settings.

Edit resolv.conf.

#+BEGIN_SRC: bash
nano /etc/resolv.conf
#+END_SRC

It should look something like the following:

#+BEGIN_SRC: bash
domain localdomain
search localdomain
nameserver 213.73.91.35
nameserver 85.214.20.141
#+END_SRC

CTRL-o followed by ENTER to save, then CTRL-x to exit.

Now disconnect the BBB from your computer and plug it into the router. You'll need an ethernet patch cable and you may also need a 5V/1A power supply for the BBB.

If you go to the web administration screen for your internet router (often it's on 192.168.2.1 or 192.168.1.254) then after a few minutes you should see the BBB appear on the network.  It's name will be "arm".

** Add a user

Ssh back in to the BBB and login as root.  In this example the BBB's IP address is 192.168.1.60.

#+BEGIN_SRC: bash
ssh-keygen -f "/home/myusername/.ssh/known_hosts" -R 192.168.1.60
ssh debian@192.168.1.60
su
#+END_SRC

Then make a new user.  It's a bad idea to add users to the sudo group, because that then means that an attacker potentially only needs to know one password in order to get administrator access to the system.  With no sudoers an attacker needs to know, or be able to obtain, two separate passwords to be able to really compromise the system.

#+BEGIN_SRC: bash
adduser myusername
#+END_SRC

Exit from the ssh login by typing "exit" a couple of times, then ssh back in as the new user.  Make sure you use a difficult to guess password/phrase, or ideally a randomly generated password used together with a password manager such as KeepassX.

Remove the default debian user.

#+BEGIN_SRC: bash
userdel -r debian
#+END_SRC

** Text editor

For an editor which is less erratic than vi when used within a remote console such as Terminator.

#+BEGIN_SRC: bash
apt-get update
apt-get install emacs
#+END_SRC

Some basic Emacs keys which will be useful to new users are:

| Load a file | CTRL-x CTRL-f |
| Save        | CTRL-x CTRL-s |
| Exit        | CTRL-x CTRL-c |

** Enable backports

To enable some newer packages add backports to the repositories.

#+BEGIN_SRC: bash
echo "deb http://ftp.us.debian.org/debian wheezy-backports main" >> /etc/apt/sources.list
apt-get update
apt-get dist-upgrade
#+END_SRC

** Configure your location/language

#+BEGIN_SRC: bash
dpkg-reconfigure locales
apt-get install keyboard-configuration
#+END_SRC

You may need to reboot for this to take effect.  To verify the change.

#+BEGIN_SRC: bash
locale -a
#+END_SRC

Set your time zone with:

#+BEGIN_SRC: bash
tzselect
#+END_SRC

For example, for British time:

#+BEGIN_SRC: bash
TZ='Europe/London'; export TZ
echo "TZ='Europe/London'; export TZ" >> ~/.bashrc
echo "TZ='Europe/London'; export TZ" >> /home/myusername/.bashrc
#+END_SRC

** Upgrade the kernel
Using a more recent kernel should improve stability of the system and also allow it to make use of hardware random number generation, which improves the overall security.  Please note that this kernel is specific to the BBB, so if you're using a Raspberry Pi, Cubieboard or other SBC then look elsewhere on the web for information about upgrading the kernel.

#+BEGIN_SRC: bash
cd /tmp
wget http://freedombone.uk.to/kernel-3.14.tar.gz
#+END_SRC

Verify it.

#+BEGIN_SRC: bash
sha256sum kernel-3.14.tar.gz
c0f76ace9ae149ffbba785dd83ee1f1b7e2d5cdcd41120c8cb3fa4aea2612753
#+END_SRC

Then extract and install it.

#+BEGIN_SRC: bash
mkdir kernel-3.14
cd kernel-3.14
tar -xzvf ../kernel-3.14.tar.gz
sh install-me.sh
reboot
#+END_SRC

After the system has rebooted you can ssh back unto it and log in as the root user.  You can check that the kernel version has changed with the command:

#+BEGIN_SRC: bash
uname -mrs
#+END_SRC

** Random number generation

#+BEGIN_VERSE
/Near as I can tell, the answer on what has been requested is everything: deliberate weakenings of encryption algorithms, deliberate weakenings of random number generations, copies of master keys, encryption of the session key with an NSA-specific key … everything./

-- Bruce Schneier, on the 2013 leaked NSA documents
#+END_VERSE

The security of encryption depends upon the randomness of the random source used on your system.  If it isn't very random then it may be far more vulnerable to cryptanalysis, and it's known that in the past some dubious agencies have encouraged the use of flawed random number generators to assist with their prurient activities.  Randomness - typically referred to as /entropy/ - is often gathered from factors such as the timing of key presses or mouse movements, but since the BBB won't have such devices plugged into it this reduces the amount of entropy available.

*** On the Beaglebone Black
Computers can't really generate truly random numbers by themselves, since they're deterministic and so operate in a highly predictable manner.  Fortunately, the BBB has an onboard hardware random number generator, which is a physical process which behaves randomly and which can then be read into the computer and stored for later use in encryption algorithms.

Information on exactly how the hardware random number generator on the Beaglebone AM335x CPU works [[http://e2e.ti.com/support/arm/sitara_arm/f/791/t/292794.aspx][seems hard to come by]], but we can later use some software to verify that it does indeed produce random numbers and hasn't been deliberately weakened.

If you are using a Beaglebone and have updated the kernel then install:

#+BEGIN_SRC: bash
apt-get install rng-tools
emacs /etc/default/rng-tools
#+END_SRC

Uncomment *HRNGDEVICE=/dev/hwrng*, save and exit then restart the daemon.

#+BEGIN_SRC: bash
service rng-tools restart
#+END_SRC

Your BBB will now use hardware to generate random numbers.

*** On other Single Board Computers
If you are not using a Beaglebone (a Cubieboard for example), or if you didn't update the kernel, then you can still improve the random number generation by installing:

#+BEGIN_SRC: bash
apt-get install haveged
#+END_SRC

*** Verifying random number quality

#+BEGIN_VERSE
/Living in a surveillance state is exactly like being guilty until proven guilty./

-- Mohammad Tarakiyee
#+END_VERSE

You can check how much randomness (entropy) is available with:

#+BEGIN_SRC: bash
cat /proc/sys/kernel/random/entropy_avail
#+END_SRC

Ideally it should be in the range 1000-4096.  If it is persistently below 500 then there may be a problem with your system which could make it less secure.

To verify that random number generation is good on the BBB run:

#+BEGIN_SRC: bash
cat /dev/hwrng | rngtest -c 1000
#+END_SRC

You should see something like this, with zero or a small number of failures:

#+BEGIN_SRC: bash
rngtest: starting FIPS tests...
rngtest: bits received from input: 20000032
rngtest: FIPS 140-2 successes: 1000
rngtest: FIPS 140-2 failures: 0
rngtest: FIPS 140-2(2001-10-10) Monobit: 0
rngtest: FIPS 140-2(2001-10-10) Poker: 0
rngtest: FIPS 140-2(2001-10-10) Runs: 0
rngtest: FIPS 140-2(2001-10-10) Long run: 0
rngtest: FIPS 140-2(2001-10-10) Continuous run: 0
rngtest: input channel speed: (min=3.104; avg=26.015; max=18.626)Gibits/s
rngtest: FIPS tests speed: (min=160.281; avg=165.696; max=168.792)Mibits/s
rngtest: Program run time: 115987 microseconds
#+END_SRC
*** Cryptotronix Hashlet
#+BEGIN_VERSE
/One must acknowledge with cryptography no amount of violence will ever solve a math problem./

-- Jacob Appelbaum
#+END_VERSE

An optional extra is the [[http://cryptotronix.com/products/hashlet/][Cryptotronix Hashlet]] which also has hardware random number generation capability via the [[./Atmel-8740-CryptoAuth-ATSHA204-Datasheet.pdf][Atmel ATSHA204]] chip.

Install the hashlet [[./images/hashlet_installed.jpg][like this]] on the BBB, then install some dependencies.

#+BEGIN_SRC: bash
apt-get install git build-essential libgcrypt11-dev texinfo
#+END_SRC

Download the source code.

#+BEGIN_SRC: bash
cd /tmp
git clone https://github.com/bashrc/hashlet.git
#+END_SRC

Now install the driver.

#+BEGIN_SRC: bash
cd hashlet
chmod o+rw /dev/i2c*
./autogen.sh
make check
make install
#+END_SRC

To check the initial state of the device:

#+BEGIN_SRC: bash
hashlet --bus=/dev/i2c-2 state
#+END_SRC

It should return the message "/Factory/".  This is intended to provide an indication that the hardware hasn't been tampered with by [[https://en.wikipedia.org/wiki/Tailored_Access_Operations][TAO]] or other shady outfits in transit.  If /i2c-2/ fails then try /i2c-1/ or /i2c-0/.

#+BEGIN_SRC: bash
hashlet --bus=/dev/i2c-2 personalize
#+END_SRC

Nothing should be returned by this command, but a file called ~/.hashlet will be generated which is the private key of the device.  This personalization process is a one-time operation which physically alters the hardware, so it would not be trivial to reset the device back to "Factory" again.  To make sure it's only accessible by the root user:

#+BEGIN_SRC: bash
chmod 400 ~/.hashlet
#+END_SRC

Now create a daemon which will create a random number generator device */dev/hashletrng*.

#+BEGIN_SRC: bash
emacs /usr/bin/hashletd
#+END_SRC

#+BEGIN_SRC: bash
#!/bin/sh

PATH='/usr/local/sbin:/usr/local/bin:/usr/bin:/sbin:/usr/sbin:/bin'
I2CBUS=2
BYTES=32
DEVICE=/dev/hashletrng

# create a device
if [ ! -e ${DEVICE} ]; then
  chmod o+rw /dev/i2c*
  mknod ${DEVICE} p
fi

while :
do
hashlet --bus=/dev/i2c-${I2CBUS} --Bytes ${BYTES} random-bytes > ${DEVICE}
done
#+END_SRC

Save and exit.  Now create an init script to run it.

#+BEGIN_SRC: bash
emacs /etc/init.d/hashlet
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
#!/bin/bash

# /etc/init.d/hashlet

### BEGIN INIT INFO
# Provides:          hashlet
# Required-Start:    $remote_fs $syslog
# Required-Stop:     $remote_fs $syslog
# Default-Start:     2 3 4 5
# Default-Stop:      0 1 6
# Short-Description: hashlet
# Description:       Creates a random number generator device
### END INIT INFO

# Author: Bob Mottram <bob@robotics.uk.to>

#Settings
SERVICE='hashlet'
LOGFILE='/dev/null'
COMMAND="/usr/bin/hashletd"
USERNAME='root'
NICELEVEL=19
HISTORY=1024
INVOCATION="nice -n ${NICELEVEL} ${COMMAND}"
PATH='/usr/local/sbin:/usr/local/bin:/usr/bin:/sbin:/usr/sbin:/bin'

hashlet_start() {
echo "Starting $SERVICE..."
su --command "screen -h ${HISTORY} -dmS ${SERVICE} ${INVOCATION}" $USERNAME
}


hashlet_stop() {
echo "Stopping $SERVICE"
su --command "screen -p 0 -S ${SERVICE} -X stuff "'^C'"" $USERNAME
}


#Start-Stop here
case "$1" in
  start)
    hashlet_start
    ;;
  stop)
    hashlet_stop
    ;;
  restart)
    hashlet_stop
    sleep 10s
    hashlet_start
    ;;
    *)
  echo "Usage: $0 {start|stop|restart}"
  exit 1
  ;;
esac

exit 0
#+END_SRC

Save and exit, then start the daemon.

#+BEGIN_SRC: bash
chmod +x /usr/bin/hashletd
chmod +x /etc/init.d/hashlet
update-rc.d hashlet defaults
service hashlet start
#+END_SRC

Then to obtain some random bytes:

#+BEGIN_SRC: bash
cat /dev/hashletrng
#+END_SRC

The rate of entropy generation by the Hashlet seems very slow compared to */dev/hwrng*, and this is most likely because of the I2C interface.  So it's probably a good idea to keep hwrng as the main random source and only use the Hashlet's random number generator for any ancillary stuff.

** Alter ssh configuration

Altering the ssh configuration will make it a little more secure than the standard Debian settings.

#+BEGIN_SRC: bash
emacs /etc/ssh/sshd_config
#+END_SRC

Check the following values:

#+BEGIN_SRC: bash
PermitRootLogin no
X11Forwarding no
ServerKeyBits 4096
Protocol 2
PermitEmptyPasswords no
StrictModes yes
TCPKeepAlive no
#+END_SRC

Append the following:

#+BEGIN_SRC: bash
ClientAliveInterval 60
ClientAliveCountMax 3
Ciphers aes256-ctr,aes192-ctr,aes128-ctr
MACs hmac-sha2-512,hmac-sha2-256,hmac-ripemd160
KexAlgorithms diffie-hellman-group-exchange-sha256,diffie-hellman-group14-sha1,diffie-hellman-group-exchange-sha1
#+END_SRC

CTRL-x CTRL-s to save, then CTRL-x CTRL-c to exit.  Now clear out any pre-existing host keys and reconfigure the ssh server.

#+BEGIN_SRC: bash
rm /etc/ssh/ssh_host_*
dpkg-reconfigure openssh-server
service ssh restart
#+END_SRC

To test the new settings log out by typing "exit" a couple of times, then log back in again with:

#+BEGIN_SRC: bash
ssh -vvv myusername@192.168.1.60
#+END_SRC

and check that some number of bits are set within a 4096 bit sized key:

#+BEGIN_SRC: bash
debug2: bits set: */4096
#+END_SRC

** Getting onto the web
Create a subdomain on [[http://freedns.afraid.org][freeDNS]].  You may need to click on "/subdomains/" a couple of times.  FreeDNS is preferred because it is one of the few domain name providers which supports genuinely free (as in beer) accounts.  So if your budget is tiny or non-existent you can still participate as a first class citizen of the internet.  If you do have money to spend there is also a premium option.

Select "/dynamic DNS/" then click "/quick cron example/"

An example would look like:

#+BEGIN_SRC: bash
4,14,24,34,44,54 * * * * root sleep 29 ; /usr/bin/timeout 200 wget -O - https://free\ dns.afraid.org/dynamic/update.php?ABCKDNRCLFHENSLKNFEGSBFLFF== >> /dev/null 2>&1 &
#+END_SRC

It's important to make sure that you change the *http* to *https*, since this will help to prevent a potential attacker from hijacking your site and redirecting it to a fake version for the purposes of obtaining your login details.

Edit */etc/crontab* and append that to the top of the file, underneath the heading line which looks like this:

#+BEGIN_SRC: bash
# m h dom mon dow user	command
#+END_SRC

In general the most frequently run crontab entries should be at the top.  Then save and exit.

Via your router's firewall settings you should now open port 22 (secure shell).  This will allow you to ssh into your BBB from any location - not just your own local network.

The freeDNS subdomain which you just created will hereafter just be refered to as "/your domain name/".

If you have multiple freedns subdomains then you may want to rationalise that a little within */etc/crontab*.  Rather than listing them all individually create a script:

#+BEGIN_SRC: bash
emacs /usr/bin/dynamicdns
#+END_SRC

Add however many freedns subdomains you have.

#+BEGIN_SRC: bash
#!/bin/bash
# subdomain name 1
wget -O - https://freedns.afraid.org/dynamic/update.php?<subdomain code 1>== >> /dev/null 2>&1
# subdomain name 2
wget -O - https://freedns.afraid.org/dynamic/update.php?<subdomain code 2>== >> /dev/null 2>&1
...
#+END_SRC

Save and exit, then make the script runnable and only readable by the root user.

#+BEGIN_SRC: bash
chmod 600 /usr/bin/dynamicdns
chmod +x /usr/bin/dynamicdns
#+END_SRC

Then within */etc/crontab*

#+BEGIN_SRC: bash
emacs /etc/crontab
#+END_SRC

You can replace the multiple freedns entries with a single line:

#+BEGIN_SRC: bash
*/10 * * * * root /usr/bin/timeout 200 /usr/bin/dynamicdns
#+END_SRC

Then save and exit and restart the cron daemon.

#+BEGIN_SRC: bash
service cron restart
#+END_SRC

If you want to know what a typical /crontab/ file might look like then see the [[Example crontab file]]

** Set the host name

#+BEGIN_SRC: bash
emacs /etc/hostname
#+END_SRC

CTRL-x CTRL-s to save, then CTRL-x CTRL-c to exit.

Also issue the command, replacing /mydomainname.com/ with your domain name.

#+BEGIN_SRC: bash
hostname mydomainname.com
#+END_SRC

You may also need to assign the same hostname separately via your router's web interface.

#+BEGIN_SRC: bash
emacs /etc/hosts
#+END_SRC

Append the following, replacing /mydomainname.com/ with your domain name.

#+BEGIN_SRC: bash
127.0.1.1       mydomainname.com
#+END_SRC

If you then run the command:

#+BEGIN_SRC: bash
hostname -f
#+END_SRC

it should return your domain name.

** Install time synchronisation

#+BEGIN_VERSE
/You may delay, but time will not./

-- Benjamin Franklin
#+END_VERSE

It's convenient to have the clock on your server automatically synchronised with other servers on the internet so that you don't need to set the clock manually.

First install some prerequisites.

#+BEGIN_SRC: bash
apt-get install build-essential automake git
#+END_SRC

Now download and install tlsdate.

#+BEGIN_SRC: bash
cd /tmp
git clone https://github.com/ioerror/tlsdate.git
cd tlsdate
./autogen.sh
./configure
make
make install
#+END_SRC

Create an init script.

#+BEGIN_SRC: bash
emacs /etc/init.d/tlsdated
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
#!/bin/sh
### BEGIN INIT INFO
# Provides:          tlsdate
# Required-Start:    $network $local_fs $remote_fs
# Required-Stop:     $local_fs $remote_fs
# Default-Start:     2 3 4 5
# Default-Stop:      0 1 6
# Short-Description: secure parasitic rdate replacement
# Description:       tlsdate sets the local clock by securely connecting with
#                    TLS to remote servers and extracting the remote time out
#                    of the secure handshake. Unlike ntpdate, tlsdate uses
#                    TCP, for instance connecting to a remote HTTPS or TLS
#                    enabled service, and provides some protection against
#                    adversaries that try to feed you malicious time
#                    information.
#
### END INIT INFO

# Author: Jacob Appelbaum <jacob@appelbaum.net>

# PATH should only include /usr/* if it runs after the mountnfs.sh script
PATH=/sbin:/usr/sbin:/bin:/usr/bin:/usr/local/sbin
DESC="secure parasitic rdate replacement daemon"
NAME=tlsdated
DAEMON=/usr/local/sbin/tlsdated
DAEMON_ARGS=""
PIDFILE=/var/run/$NAME.pid
SCRIPTNAME=/etc/init.d/$NAME

# Exit if the package is not installed
[ -x $DAEMON ] || exit 0

# Read configuration variable file if it is present
[ -r /etc/default/$NAME ] && . /etc/default/$NAME

# Load the VERBOSE setting and other rcS variables
. /lib/init/vars.sh

# Define LSB log_* functions.
# Depend on lsb-base (>= 3.0-6) to ensure that this file is present.
. /lib/lsb/init-functions

#
# Function that starts the daemon/service
#
do_start()
{
    # Return
    #   0 if daemon has been started
    #   1 if daemon was already running
    #   2 if daemon could not be started
    start-stop-daemon --background --start --quiet --pidfile $PIDFILE \
		--exec $DAEMON --test > /dev/null \
        || return 1
    start-stop-daemon --background --start --quiet --pidfile $PIDFILE \
		--exec $DAEMON -- \
        $DAEMON_ARGS \
        || return 2
    # Add code here, if necessary, that waits for the process to be ready
    # to handle requests from services started subsequently which depend
    # on this one.  As a last resort, sleep for some time.
}

#
# Function that stops the daemon/service
#
do_stop()
{
    # Return
    #   0 if daemon has been stopped
    #   1 if daemon was already stopped
    #   2 if daemon could not be stopped
    #   other if a failure occurred
    start-stop-daemon --stop --quiet --retry=TERM/5/KILL/1 --pidfile $PIDFILE \
		--name $NAME
        RETVAL="$?"
        [ "$RETVAL" = 2 ] && return 2
    # Wait for children to finish too if this is a daemon that forks
    # and if the daemon is only ever run from this initscript.
    # If the above conditions are not satisfied then add some other code
    # that waits for the process to drop all resources that could be
    # needed by services started subsequently.  A last resort is to
    # sleep for some time.
    start-stop-daemon --stop --quiet --oknodo --retry=0/5/KILL/5 --exec $DAEMON
        [ "$?" = 2 ] && return 2
        # Many daemons don't delete their pidfiles when they exit.
        rm -f $PIDFILE
        return "$RETVAL"
}

#
# Function that sends a SIGHUP to the daemon/service
#
do_reload() {
    #
    # If the daemon can reload its configuration without
    # restarting (for example, when it is sent a SIGHUP),
    # then implement that here.
    #
    start-stop-daemon --stop --signal 1 --quiet --pidfile $PIDFILE --name $NAME
    return 0
}

case "$1" in
  start)
    [ "$VERBOSE" != no ] && log_daemon_msg "Starting $DESC " "$NAME"
    do_start
    case "$?" in
         0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
              2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
              esac
  ;;
  stop)
  [ "$VERBOSE" != no ] && log_daemon_msg "Stopping $DESC" "$NAME"
  do_stop
  case "$?" in
       0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
            2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
            esac
            ;;
  status)
       status_of_proc "$DAEMON" "$NAME" && exit 0 || exit $?
       ;;
  #reload|force-reload)
  #
  # If do_reload() is not implemented then leave this commented out
  # and leave 'force-reload' as an alias for 'restart'.
  #
  #log_daemon_msg "Reloading $DESC" "$NAME"
  #do_reload
  #log_end_msg $?
  #;;
  restart|force-reload)
  #
  # If the "reload" option is implemented then remove the
  # 'force-reload' alias
  #
  log_daemon_msg "Restarting $DESC" "$NAME"
  do_stop
  case "$?" in
    0|1)
        do_start
        case "$?" in
            0) log_end_msg 0 ;;
            1) log_end_msg 1 ;; # Old process is still running
            *) log_end_msg 1 ;; # Failed to start
            esac
        ;;
    *)
    # Failed to stop
    log_end_msg 1
    ;;
    esac
    ;;
  *)
  echo "Usage: $SCRIPTNAME {start|stop|status|restart|force-reload}" >&2
  exit 3
  ;;
esac

:
#+END_SRC

Save and exit, then start the daemon.

#+BEGIN_SRC: bash
chmod +x /etc/init.d/tlsdated
update-rc.d tlsdated defaults
service tlsdated start
#+END_SRC

** Install fail2ban

#+BEGIN_SRC: bash
apt-get install fail2ban
#+END_SRC

** Set up a firewall

#+BEGIN_VERSE
/The documents, from a PowerPoint presentation prepared for a 2012 NSA conference called SIGDEV, show that the unit known as the Joint Threat Research Intelligence Group, or JTRIG, boasted of using the DDOS attack – which it dubbed Rolling Thunder/

-- NBC News article: /War on Anonymous: British Spies Attacked Hackers, Snowden Docs Show/
#+END_VERSE

A basic firewall limits the maximum rate at which connections can be made and closes any unused ports, and this helps to defend against various kinds of DDOS attack.

#+BEGIN_SRC: bash
apt-get install portsentry
emacs /etc/portsentry/portsentry.conf
#+END_SRC

Uncomment the entry for *iptables support for Linux*

Set the following properties:

#+BEGIN_SRC: bash
TCP_PORTS="1,7,9,11,15,79,109,110,111,119,138,139,512,513,514,515,540,635,1080,1524,2000,2001,4000,4001,5742,6000,6001,6667,12345,12346,20034,27665,30303,32771,32772,32773,32774,31337,40421,40425,49724,54320"
UDP_PORTS="1,7,9,66,67,68,69,111,137,138,161,162,474,513,517,518,635,640,641,666,700,2049,31335,27444,34555,32770,32771,32772,32773,32774,31337,54321"

ADVANCED_EXCLUDE_TCP="113,139,70,80,443,143,6670,993,5060,5061,25,465,22,5222,5223,5269,5280,5281,8432,8433,8444"
ADVANCED_EXCLUDE_UDP="520,138,137,67,70,80,443,143,6670,993, 5060,5061,25,465,22,5222,5223,5269,5280,5281,8444"

SCAN_TRIGGER="2"

BLOCK_UDP="2"
BLOCK_TCP="2"
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
service portsentry restart
emacs /tmp/firewall.sh
#+END_SRC

Enter the following:

#+BEGIN_SRC: bash
#!/bin/bash

# First of all delete any existing rules.
# This means you're back to a known state:
iptables -P INPUT ACCEPT
ip6tables -P INPUT ACCEPT
iptables -F
ip6tables -F
iptables -X
ip6tables -X

# Drop any IPv6 traffic
ip6tables -A INPUT -p icmp -j DROP
ip6tables -A INPUT -p tcp -j DROP
ip6tables -A INPUT -p udp -j DROP

# Drop access to unused ports
iptables -A INPUT -p tcp --destination-port 1 -j DROP
iptables -A INPUT -p tcp --destination-port 7 -j DROP
iptables -A INPUT -p tcp --destination-port 109:111 -j DROP
iptables -A INPUT -p tcp --destination-port 995 -j DROP
iptables -A INPUT -p tcp --destination-port 139 -j DROP
iptables -A INPUT -p tcp --destination-port 6000:6001 -j DROP
iptables -A INPUT -p tcp --destination-port 9 -j DROP
iptables -A INPUT -p tcp --destination-port 79 -j DROP
iptables -A INPUT -p tcp --destination-port 515 -j DROP
iptables -A INPUT -p tcp --destination-port 4001 -j DROP
iptables -A INPUT -p tcp --destination-port 1524 -j DROP
iptables -A INPUT -p tcp --destination-port 1080 -j DROP
iptables -A INPUT -p tcp --destination-port 512:514 -j DROP
iptables -A INPUT -p tcp --destination-port 31337 -j DROP
iptables -A INPUT -p tcp --destination-port 2000:2001 -j DROP
iptables -A INPUT -p tcp --destination-port 12345 -j DROP
iptables -A INPUT -p tcp --destination-port 32771:32774 -j DROP
iptables -A INPUT -p tcp --destination-port 4000 -j DROP
iptables -A INPUT -p tcp --destination-port 119 -j DROP
iptables -A INPUT -p tcp --destination-port 137 -j DROP
iptables -A INPUT -p udp --destination-port 1 -j DROP
iptables -A INPUT -p udp --destination-port 7 -j DROP
iptables -A INPUT -p udp --destination-port 109:111 -j DROP
iptables -A INPUT -p udp --destination-port 995 -j DROP
iptables -A INPUT -p udp --destination-port 139 -j DROP
iptables -A INPUT -p udp --destination-port 6000:6001 -j DROP
iptables -A INPUT -p udp --destination-port 9 -j DROP
iptables -A INPUT -p udp --destination-port 79 -j DROP
iptables -A INPUT -p udp --destination-port 515 -j DROP
iptables -A INPUT -p udp --destination-port 4001 -j DROP
iptables -A INPUT -p udp --destination-port 1524 -j DROP
iptables -A INPUT -p udp --destination-port 1080 -j DROP
iptables -A INPUT -p udp --destination-port 512:514 -j DROP
iptables -A INPUT -p udp --destination-port 31337 -j DROP
iptables -A INPUT -p udp --destination-port 2000:2001 -j DROP
iptables -A INPUT -p udp --destination-port 12345 -j DROP
iptables -A INPUT -p udp --destination-port 32771:32774 -j DROP
iptables -A INPUT -p udp --destination-port 4000 -j DROP
iptables -A INPUT -p udp --destination-port 119 -j DROP
iptables -A INPUT -p udp --destination-port 137 -j DROP
iptables -A INPUT -p udp --destination-port 8432 -j DROP
iptables -A INPUT -p udp --destination-port 8433 -j DROP

# Make sure NEW incoming tcp connections are SYN packets
iptables -A INPUT -p tcp ! --syn -m state --state NEW -j DROP

# Drop packets with incoming fragments
iptables -A INPUT -f -j DROP

# Incoming malformed XMAS packets drop them
iptables -A INPUT -p tcp --tcp-flags ALL ALL -j DROP
iptables -A INPUT -p tcp --tcp-flags ALL FIN,PSH,URG -j DROP
iptables -A INPUT -p tcp --tcp-flags ALL SYN,RST,ACK,FIN,URG -j DROP

# Incoming malformed NULL packets:
iptables -A INPUT -p tcp --tcp-flags ALL NONE -j DROP

# Drop UDP to used ports
iptables -A INPUT -p udp --match multiport --dports 70,80,443,143,6670,993,5060,5061,25 -j DROP
iptables -A INPUT -p udp --match multiport --dports 465,22,5222,5223,5269,5280,5281,8444 -j DROP

# Limit ssh logins
iptables -A INPUT -p tcp --dport 22 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT

# Limit web connections
iptables -A INPUT -p tcp --dport 80 -j ACCEPT
iptables -A INPUT -p tcp --dport 443 -j ACCEPT

# Limit number of XMPP connections
iptables -A INPUT -p tcp --match multiport --dports 5222:5223,5269,5280:5281 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT

# Limit IRC connections
iptables -A INPUT -p tcp --dport 6666:6670 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT

# Limit gopher connections
iptables -A INPUT -p tcp --dport 70 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT

# Limit IMAP connections
iptables -A INPUT -p tcp --dport 143 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT
iptables -A INPUT -p tcp --dport 993 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT

# Limit SIP connections
iptables -A INPUT -p tcp --dport 5060:5061 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT

# Limit SMTP/SMTPS connections
iptables -A INPUT -p tcp --dport 25 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT
iptables -A INPUT -p tcp --dport 465 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT

# Limit Bitmessage connections
iptables -A INPUT -p tcp --dport 8444 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT

# Limit Convergence notary
iptables -A INPUT -p tcp --dport 8432:8433 -m limit --limit 3/minute --limit-burst 1 -j ACCEPT

# Limit the number of incoming tcp connections
# Interface 0 incoming syn-flood protection
iptables -N syn_flood
iptables -A INPUT -p tcp --syn -j syn_flood
iptables -A syn_flood -m limit --limit 1/s --limit-burst 3 -j RETURN
iptables -A syn_flood -j DROP

# Limiting the incoming icmp ping request:
#iptables -A INPUT -p icmp -m limit --limit  1/s --limit-burst 1 -j ACCEPT
#iptables -A INPUT -p icmp -m limit --limit 1/s --limit-burst 1 -j LOG --log-prefix PING-DROP:
iptables -A INPUT -p icmp -j DROP
#iptables -A OUTPUT -p icmp -j ACCEPT

# Save the settings
iptables-save > /etc/firewall.conf
ip6tables-save > /etc/firewall6.conf
printf '#!/bin/sh\n' > /etc/network/if-up.d/iptables
printf 'iptables-restore < /etc/firewall.conf\n' >> /etc/network/if-up.d/iptables
printf 'ip6tables-restore < /etc/firewall6.conf\n' >> /etc/network/if-up.d/iptables
chmod +x /etc/network/if-up.d/iptables
#+END_SRC

Save and exit.

Note that this will disable IP version 6.  At the time of writing it is expected that the average internet user is running on IP version 4.

#+BEGIN_SRC: bash
chmod +x /tmp/firewall.sh
. /tmp/firewall.sh
rm /tmp/firewall.sh
#+END_SRC

Also disable ping.  This may be inconvenient to some extent, but it seems common for malicious systems, including but not limited to the [[http://www.nbcnews.com/news/investigations/snowden-docs-british-spies-used-sex-dirty-tricks-n23091][JTRIG "EFFECTS" team]], to try to disable the machine by flooding it with pings.  These days there seems to be not much difference between "cybercrime" and nefarious state-sponsored internet activities.

#+BEGIN_SRC: bash
emacs /etc/sysctl.conf
#+END_SRC

Uncomment or change the following:

#+BEGIN_SRC: bash
net.ipv4.tcp_syncookies = 1
net.ipv4.conf.all.accept_redirects = 0
net.ipv6.conf.all.accept_redirects = 0
net.ipv4.conf.all.send_redirects = 0
net.ipv4.conf.all.accept_source_route = 0
net.ipv6.conf.all.accept_source_route = 0
net.ipv4.conf.default.rp_filter=1
net.ipv4.conf.all.rp_filter=1
net.ipv4.ip_forward=0
net.ipv6.conf.all.forwarding=0
#+END_SRC

And append the following:

#+BEGIN_SRC: bash
# ignore pings
net.ipv4.icmp_echo_ignore_all = 1
net.ipv6.icmp_echo_ignore_all = 1

# disable ipv6
net.ipv6.conf.all.disable_ipv6 = 1

net.ipv4.tcp_synack_retries = 2
net.ipv4.tcp_syn_retries = 1

# keepalive
net.ipv4.tcp_keepalive_probes = 9
net.ipv4.tcp_keepalive_intvl = 75
net.ipv4.tcp_keepalive_time = 7200
#+END_SRC

Save and exit.  It may be a good idea to reboot at this point and then log back into the BBB using ssh.  You can do a safe reboot of the system by typing:

#+BEGIN_SRC: bash
reboot
#+END_SRC

After reboot and logging back in to the root account via /ssh/ you can verify that the firewall rules were restored correctly with:

#+BEGIN_SRC: bash
iptables -L
#+END_SRC

and

#+BEGIN_SRC: bash
ip6tables -L
#+END_SRC

** Install Email

#+BEGIN_VERSE
/If you knew what I know about email, you might not use it/

-- Ladar Levison
#+END_VERSE

Email is not very secure, but its usefulness and ubiquity mean that it's likely to continue as a primary communications method for many years to come.  You can encrypt the contents of email using PGP/GPG, but very few people do that and even for those that do the metadata (the From/To/CC/BCC) is always transmitted in the clear as a fundamental aspect of the protocol, allowing an attacker to easily construct detailed models of people's social network and life patterns even without knowing the content.

Exim4 seems much easier to install and configure than Postfix.

#+BEGIN_SRC: bash
service postfix stop
apt-get remove postfix
aptitude install exim4 sasl2-bin swaks libnet-ssleay-perl procmail
#+END_SRC

You will be prompted to remove postfix.  Say yes and yes again.

#+BEGIN_SRC: bash
dpkg-reconfigure exim4-config
#+END_SRC

Settings as follows:

#+BEGIN_SRC: bash
internet site

System mail name: mydomainname.com

IP addresses to listen on: blank

Destinations: mydomainname.com

Domains to relay mail: blank

Smarthost Relay: 192.168.1.0/24 (the range of addresses on your LAN)

Dial on demand = no

Maildir format in home directory

Split configuration = no

Root and postmaster: root email
#+END_SRC

To test the installation:

#+BEGIN_SRC: bash
telnet 192.168.1.60 25
ehlo xxx
quit
#+END_SRC

#+BEGIN_SRC: bash
emacs /etc/default/saslauthd
#+END_SRC

set START=yes then save and exit.

#+BEGIN_SRC: bash
/etc/init.d/saslauthd start
emacs exim-gencert
#+END_SRC

#+BEGIN_SRC: bash
#!/bin/sh -e

if [ -n "$EX4DEBUG" ]; then
  echo "now debugging $0 $@"
  set -x
fi

DIR=/etc/exim4
CERT=$DIR/exim.crt
KEY=$DIR/exim.key

# This exim binary was built with GnuTLS which does not support dhparams
# from a file. See /usr/share/doc/exim4-base/README.Debian.gz
#DH=$DIR/exim.dhparam

if ! which openssl > /dev/null ;then
    echo "$0: openssl is not installed, exiting" 1>&2
    exit 1
fi

# valid for ten years
DAYS=3650

if [ "$1" != "--force" ] && [ -f $CERT ] && [ -f $KEY ]; then
  echo "[*] $CERT and $KEY exists!"
  echo "    Use \"$0 --force\" to force generation!"
  exit 0
fi

if [ "$1" = "--force" ]; then
  shift
fi

#SSLEAY=/tmp/exim.ssleay.$$.cnf
SSLEAY="$(tempfile -m600 -pexi)"
cat > $SSLEAY <<EOM
RANDFILE = $HOME/.rnd
[ req ]
default_bits = 4096
default_keyfile = exim.key
distinguished_name = req_distinguished_name
[ req_distinguished_name ]
countryName = Country Code (2 letters)
countryName_default = GB
countryName_min = 2
countryName_max = 2
stateOrProvinceName = State or Province Name (full name)
localityName = Locality Name (eg, city)
organizationName = Organization Name (eg, company; recommended)
organizationName_max = 64
organizationalUnitName = Organizational Unit Name (eg, section)
organizationalUnitName_max = 64
commonName = Server name (eg. ssl.domain.tld; required!!!)
commonName_max = 64
emailAddress = Email Address
emailAddress_max = 40
EOM

echo "[*] Creating a self signed SSL certificate for Exim!"
echo "    This may be sufficient to establish encrypted connections but for"
echo "    secure identification you need to buy a real certificate!"
echo "    "
echo "    Please enter the hostname of your MTA at the Common Name (CN) prompt!"
echo "    "

openssl req -config $SSLEAY -x509 -sha256 -newkey rsa:4096 -keyout $KEY -out $CERT -days $DAYS -nodes
#see README.Debian.gz*# openssl dhparam -check -text -5 512 -out $DH
rm -f $SSLEAY

chown root:Debian-exim $KEY $CERT $DH
chmod 640 $KEY $CERT $DH

echo "[*] Done generating self signed certificates for exim!"
echo "    Refer to the documentation and example configuration files"
echo "    over at /usr/share/doc/exim4-base/ for an idea on how to enable TLS"
echo "    support in your mail transfer agent."
#+END_SRC

Save and exit

#+BEGIN_SRC: bash
chmod +x exim-gencert
./exim-gencert
#+END_SRC

This will generate the certificate used for email authentication.  You will be asked for various details, the most important of which is the server name, which should be your domain name.

#+BEGIN_SRC: bash
emacs /etc/exim4/exim4.conf.template
#+END_SRC

Append the following:

#+BEGIN_SRC: bash
login_saslauthd_server:
  driver = plaintext
  public_name = LOGIN
  server_prompts = "Username:: : Password::"
  # don't send system passwords over unencrypted connections
  server_condition = ${if saslauthd{{$auth1}{$auth2}}{1}{0}}
  server_set_id = $auth1
 .ifndef AUTH_SERVER_ALLOW_NOTLS_PASSWORDS
 server_advertise_condition = ${if eq{$tls_cipher}{}{}{*}}
 .endif
#+END_SRC

Search for the line *.ifdef MAIN_HARDCODE_PRIMARY_HOSTNAME* and above it insert the line:

#+BEGIN_SRC: bash
MAIN_HARDCODE_PRIMARY_HOSTNAME = mydomainname.com
#+END_SRC

Add the line:

#+BEGIN_SRC: bash
MAIN_TLS_ENABLE = true
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/default/exim4
change SMTPLISTENEROPTIONS to:
SMTPLISTENEROPTIONS='-oX 465:25 -oP /var/run/exim4/exim.pid'
#+END_SRC

save and exit

#+BEGIN_SRC: bash
emacs /etc/exim4/exim4.conf.template
under the section "main/03_exim4-config_tlsoptions"
Add the following:
tls_on_connect_ports=465
#+END_SRC

save and exit

#+BEGIN_SRC: bash
adduser myusername sasl
addgroup Debian-exim sasl
/etc/init.d/exim4 restart
mkdir -m 700 /etc/skel/Maildir
mkdir -m 700 /etc/skel/Maildir/Sent
mkdir -m 700 /etc/skel/Maildir/Sent/tmp
mkdir -m 700 /etc/skel/Maildir/Sent/cur
mkdir -m 700 /etc/skel/Maildir/Sent/new
mkdir -m 700 /etc/skel/Maildir/.learn-spam
mkdir -m 700 /etc/skel/Maildir/.learn-spam/cur
mkdir -m 700 /etc/skel/Maildir/.learn-spam/new
mkdir -m 700 /etc/skel/Maildir/.learn-spam/tmp
mkdir -m 700 /etc/skel/Maildir/.learn-ham
mkdir -m 700 /etc/skel/Maildir/.learn-ham/cur
mkdir -m 700 /etc/skel/Maildir/.learn-ham/new
mkdir -m 700 /etc/skel/Maildir/.learn-ham/tmp
ln -s /etc/skel/Maildir/.learn-spam /etc/skel/Maildir/spam
ln -s /etc/skel/Maildir/.learn-ham /etc/skel/Maildir/ham
#+END_SRC

If you're starting from scratch and don't already have a /Maildir/ directory in your home directory, then create one as follows:

#+BEGIN_SRC: bash
export MYUSERNAME=myusername
mkdir -m 700 /home/$MYUSERNAME/Maildir
mkdir -m 700 /home/$MYUSERNAME/Maildir/cur
mkdir -m 700 /home/$MYUSERNAME/Maildir/tmp
mkdir -m 700 /home/$MYUSERNAME/Maildir/new
mkdir -m 700 /home/$MYUSERNAME/Maildir/Sent
mkdir -m 700 /home/$MYUSERNAME/Maildir/Sent/cur
mkdir -m 700 /home/$MYUSERNAME/Maildir/Sent/tmp
mkdir -m 700 /home/$MYUSERNAME/Maildir/Sent/new
mkdir -m 700 /home/$MYUSERNAME/Maildir/.learn-spam
mkdir -m 700 /home/$MYUSERNAME/Maildir/.learn-spam/cur
mkdir -m 700 /home/$MYUSERNAME/Maildir/.learn-spam/new
mkdir -m 700 /home/$MYUSERNAME/Maildir/.learn-spam/tmp
mkdir -m 700 /home/$MYUSERNAME/Maildir/.learn-ham
mkdir -m 700 /home/$MYUSERNAME/Maildir/.learn-ham/cur
mkdir -m 700 /home/$MYUSERNAME/Maildir/.learn-ham/new
mkdir -m 700 /home/$MYUSERNAME/Maildir/.learn-ham/tmp
ln -s /home/$MYUSERNAME/Maildir/.learn-spam /home/$MYUSERNAME/Maildir/spam
ln -s /home/$MYUSERNAME/Maildir/.learn-ham /home/$MYUSERNAME/Maildir/ham
chown -R $MYUSERNAME:$MYUSERNAME /home/$MYUSERNAME/Maildir
#+END_SRC

** Spam filtering

#+BEGIN_SRC: bash
apt-get install spamassassin exim4-daemon-heavy
emacs /etc/default/spamassassin
#+END_SRC

Set ENABLED=1 then save and exit.

#+BEGIN_SRC: bash
emacs /etc/exim4/exim4.conf.template
#+END_SRC

uncomment or change according to your configuration

#+BEGIN_SRC: bash
# For spam scanning, there is a similar option that defines the interface to
# SpamAssassin. You do not need to set this if you are using the default, which
# is shown in this commented example. As for virus scanning, you must also
# modify the acl_check_data access control list to enable spam scanning.

 spamd_address = 127.0.0.1 783
#+END_SRC

add spam header in the /acl_check_data/ section:

#+BEGIN_SRC: bash
### acl/40_exim4-config_check_data
#################################

# This ACL is used after the contents of a message have been received. This
# is the ACL in which you can test a message's headers or body, and in
# particular, this is where you can invoke external virus or spam scanners.

acl_check_data:
...
...
...
# See the exim docs and the exim wiki for more suitable examples.
#
# warn
#   spam = Debian-exim:true
#   add_header = X-Spam_score: $spam_score\n\
#             X-Spam_score_int: $spam_score_int\n\
#             X-Spam_bar: $spam_bar\n\
#             X-Spam_report: $spam_report

# put headers in all messages (no matter if spam or not)
 warn  spam = nobody:true
     add_header = X-Spam-Score: $spam_score ($spam_bar)
     add_header = X-Spam-Report: $spam_report

# add second subject line with *SPAM* marker when message
# is over threshold
  warn  spam = nobody
      add_header = Subject: ***SPAM (score:$spam_score)*** $h_Subject:
#+END_SRC

Save and exit.

Then restart

#+BEGIN_SRC: bash
exit
emacs ~/.procmailrc
#+END_SRC

The text should look like the following.

#+BEGIN_SRC: sh
MAILDIR=$HOME/Maildir
DEFAULT=$MAILDIR/
LOGFILE=$HOME/log/procmail.log
LOGABSTRACT=all

# get spamassassin to check emails
:0fw: .spamassassin.lock
  * < 256000
| spamc

# strong spam are discarded
:0
  * ^X-Spam-Level: \*\*\*\*\*\*
/dev/null

# weak spam are kept just in case - clear this out every now and then
:0
  * ^X-Spam-Level: \*\*\*\*\*
.0-spam/

# otherwise, marginal spam goes here for revision
:0
  * ^X-Spam-Level: \*\*
.spam/
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
su
emacs /usr/bin/filterspam
#+END_SRC

Add the following contents:

#+BEGIN_SRC: bash
#!/bin/bash

USERNAME=$1
MAILDIR=/home/$USERNAME/Maildir/.learn-spam

if [ ! -d "$MAILDIR" ]; then
    exit
fi

for f in `ls $MAILDIR/cur`
do
    spamc -L spam < "$MAILDIR/cur/$f" > /dev/null
    rm "$MAILDIR/cur/$f"
done
for f in `ls $MAILDIR/new`
do
    spamc -L spam < "$MAILDIR/new/$f" > /dev/null
    rm "$MAILDIR/new/$f"
done
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /usr/bin/filterham
#+END_SRC

Add the following contents:

#+BEGIN_SRC: bash
#!/bin/bash

USERNAME=$1
MAILDIR=/home/$USERNAME/Maildir/.learn-ham

if [ ! -d "$MAILDIR" ]; then
    exit
fi

for f in `ls $MAILDIR/cur`
do
    spamc -L ham < "$MAILDIR/cur/$f" > /dev/null
    rm "$MAILDIR/cur/$f"
done
for f in `ls $MAILDIR/new`
do
    spamc -L ham < "$MAILDIR/new/$f" > /dev/null
    rm "$MAILDIR/new/$f"
done
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/crontab
#+END_SRC

Append the following, replacing *myusername* with your username.

#+BEGIN_SRC: bash
*/3 * * * * root /usr/bin/timeout 120 /usr/bin/filterspam myusername
*/3 * * * * root /usr/bin/timeout 120 /usr/bin/filterham myusername
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod 655 /usr/bin/filterspam /usr/bin/filterham
service spamassassin restart
service exim4 restart
service cron restart
#+END_SRC

** Install Dovecot

#+BEGIN_VERSE
/I dreamt last night that I was living in a surveillance state. I woke up and… I’m still in a surveillance state./

-- Conrad Kramer
#+END_VERSE

Install the required packages.

#+BEGIN_SRC: bash
aptitude -y install dovecot-common dovecot-imapd
#+END_SRC

Edit the configuration file.

#+BEGIN_SRC: bash
emacs /etc/dovecot/dovecot.conf
#+END_SRC

Line 26: change:

#+BEGIN_SRC: bash
listen = *
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/dovecot/conf.d/10-auth.conf
#+END_SRC

Line 9: uncomment and change (allow plain text auth)

#+BEGIN_SRC: bash
disable_plaintext_auth = no
#+END_SRC

Line 99: add:

#+BEGIN_SRC: bash
auth_mechanisms = plain login
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/dovecot/conf.d/10-mail.conf
#+END_SRC

Line 30: uncomment and add:

#+BEGIN_SRC: bash
mail_location = maildir:~/Maildir:LAYOUT=fs
#+END_SRC

Save and exit, then start the dovecot service.

#+BEGIN_SRC: bash
service dovecot restart
#+END_SRC

** Create a GPG key
#+BEGIN_VERSE
/If privacy is outlawed, only outlaws will have privacy./

-- Philip Zimmermann
#+END_VERSE

*** Initial installation

Assuming that you are logged in as root, first ensure that GPG is installed and then exit to your user account.

#+BEGIN_SRC: bash
apt-get install gnupg
exit
#+END_SRC

Now we will add some settings:

#+BEGIN_SRC: bash
mkdir ~/.gnupg
emacs ~/.gnupg/gpg.conf
#+END_SRC

The configuration should look like the following.  Of particular importance are the default preferences at the end.

#+BEGIN_SRC: bash
# Options for GnuPG
# Copyright 1998, 1999, 2000, 2001, 2002, 2003,
#           2010 Free Software Foundation, Inc.
#
# This file is free software; as a special exception the author gives
# unlimited permission to copy and/or distribute it, with or without
# modifications, as long as this notice is preserved.
#
# This file is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY, to the extent permitted by law; without even the
# implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
#
# Unless you specify which option file to use (with the command line
# option "--options filename"), GnuPG uses the file ~/.gnupg/gpg.conf
# by default.
#
# An options file can contain any long options which are available in
# GnuPG. If the first non white space character of a line is a '#',
# this line is ignored.  Empty lines are also ignored.
#
# See the man page for a list of options.

# Uncomment the following option to get rid of the copyright notice

#no-greeting

# If you have more than 1 secret key in your keyring, you may want to
# uncomment the following option and set your preferred keyid.

#default-key 621CC013

# If you do not pass a recipient to gpg, it will ask for one.  Using
# this option you can encrypt to a default key.  Key validation will
# not be done in this case.  The second form uses the default key as
# default recipient.

#default-recipient some-user-id
#default-recipient-self

# Use --encrypt-to to add the specified key as a recipient to all
# messages.  This is useful, for example, when sending mail through a
# mail client that does not automatically encrypt mail to your key.
# In the example, this option allows you to read your local copy of
# encrypted mail that you've sent to others.

#encrypt-to some-key-id

# By default GnuPG creates version 4 signatures for data files as
# specified by OpenPGP.  Some earlier (PGP 6, PGP 7) versions of PGP
# require the older version 3 signatures.  Setting this option forces
# GnuPG to create version 3 signatures.

#force-v3-sigs

# Because some mailers change lines starting with "From " to ">From "
# it is good to handle such lines in a special way when creating
# cleartext signatures; all other PGP versions do it this way too.

#no-escape-from-lines

# If you do not use the Latin-1 (ISO-8859-1) charset, you should tell
# GnuPG which is the native character set.  Please check the man page
# for supported character sets.  This character set is only used for
# metadata and not for the actual message which does not undergo any
# translation.  Note that future version of GnuPG will change to UTF-8
# as default character set.  In most cases this option is not required
# as GnuPG is able to figure out the correct charset at runtime.

#charset utf-8

# Group names may be defined like this:
#   group mynames = paige 0x12345678 joe patti
#
# Any time "mynames" is a recipient (-r or --recipient), it will be
# expanded to the names "paige", "joe", and "patti", and the key ID
# "0x12345678".  Note there is only one level of expansion - you
# cannot make an group that points to another group.  Note also that
# if there are spaces in the recipient name, this will appear as two
# recipients.  In these cases it is better to use the key ID.

#group mynames = paige 0x12345678 joe patti

# Lock the file only once for the lifetime of a process.  If you do
# not define this, the lock will be obtained and released every time
# it is needed, which is usually preferable.

#lock-once

# GnuPG can send and receive keys to and from a keyserver.  These
# servers can be HKP, email, or LDAP (if GnuPG is built with LDAP
# support).
#
# Example HKP keyserver:
#      hkp://keys.gnupg.net
#      hkp://subkeys.pgp.net
#
# Example email keyserver:
#      mailto:pgp-public-keys@keys.pgp.net
#
# Example LDAP keyservers:
#      ldap://keyserver.pgp.com
#
# Regular URL syntax applies, and you can set an alternate port
# through the usual method:
#      hkp://keyserver.example.net:22742
#
# Most users just set the name and type of their preferred keyserver.
# Note that most servers (with the notable exception of
# ldap://keyserver.pgp.com) synchronize changes with each other.  Note
# also that a single server name may actually point to multiple
# servers via DNS round-robin.  hkp://keys.gnupg.net is an example of
# such a "server", which spreads the load over a number of physical
# servers.  To see the IP address of the server actually used, you may use
# the "--keyserver-options debug".

keyserver hkp://keys.gnupg.net
#keyserver mailto:pgp-public-keys@keys.nl.pgp.net
#keyserver ldap://keyserver.pgp.com

# Common options for keyserver functions:
#
# include-disabled : when searching, include keys marked as "disabled"
#                    on the keyserver (not all keyservers support this).
#
# no-include-revoked : when searching, do not include keys marked as
#                      "revoked" on the keyserver.
#
# verbose : show more information as the keys are fetched.
#           Can be used more than once to increase the amount
#           of information shown.
#
# use-temp-files : use temporary files instead of a pipe to talk to the
#                  keyserver.  Some platforms (Win32 for one) always
#                  have this on.
#
# keep-temp-files : do not delete temporary files after using them
#                   (really only useful for debugging)
#
# http-proxy="proxy" : set the proxy to use for HTTP and HKP keyservers.
#                      This overrides the "http_proxy" environment variable,
#                      if any.
#
# auto-key-retrieve : automatically fetch keys as needed from the keyserver
#                     when verifying signatures or when importing keys that
#                     have been revoked by a revocation key that is not
#                     present on the keyring.
#
# no-include-attributes : do not include attribute IDs (aka "photo IDs")
#                         when sending keys to the keyserver.

keyserver-options auto-key-retrieve

# Display photo user IDs in key listings

# list-options show-photos

# Display photo user IDs when a signature from a key with a photo is
# verified

# verify-options show-photos

# Use this program to display photo user IDs
#
# %i is expanded to a temporary file that contains the photo.
# %I is the same as %i, but the file isn't deleted afterwards by GnuPG.
# %k is expanded to the key ID of the key.
# %K is expanded to the long OpenPGP key ID of the key.
# %t is expanded to the extension of the image (e.g. "jpg").
# %T is expanded to the MIME type of the image (e.g. "image/jpeg").
# %f is expanded to the fingerprint of the key.
# %% is %, of course.
#
# If %i or %I are not present, then the photo is supplied to the
# viewer on standard input.  If your platform supports it, standard
# input is the best way to do this as it avoids the time and effort in
# generating and then cleaning up a secure temp file.
#
# If no photo-viewer is provided, GnuPG will look for xloadimage, eog,
# or display (ImageMagick).  On Mac OS X and Windows, the default is
# to use your regular JPEG image viewer.
#
# Some other viewers:
# photo-viewer "qiv %i"
# photo-viewer "ee %i"
#
# This one saves a copy of the photo ID in your home directory:
# photo-viewer "cat > ~/photoid-for-key-%k.%t"
#
# Use your MIME handler to view photos:
# photo-viewer "metamail -q -d -b -c %T -s 'KeyID 0x%k' -f GnuPG"

# Passphrase agent
#
# We support the old experimental passphrase agent protocol as well as
# the new Assuan based one (currently available in the "newpg" package
# at ftp.gnupg.org/gcrypt/alpha/aegypten/).  To make use of the agent,
# you have to run an agent as daemon and use the option
#
# use-agent
#
# which tries to use the agent but will fallback to the regular mode
# if there is a problem connecting to the agent.  The normal way to
# locate the agent is by looking at the environment variable
# GPG_AGENT_INFO which should have been set during gpg-agent startup.
# In certain situations the use of this variable is not possible, thus
# the option
#
# --gpg-agent-info=<path>:<pid>:1
#
# may be used to override it.

# Automatic key location
#
# GnuPG can automatically locate and retrieve keys as needed using the
# auto-key-locate option.  This happens when encrypting to an email
# address (in the "user@example.com" form), and there are no
# user@example.com keys on the local keyring.  This option takes the
# following arguments, in the order they are to be tried:
#
# cert = locate a key using DNS CERT, as specified in RFC-4398.
#        GnuPG can handle both the PGP (key) and IPGP (URL + fingerprint)
#        CERT methods.
#
# pka = locate a key using DNS PKA.
#
# ldap = locate a key using the PGP Universal method of checking
#        "ldap://keys.(thedomain)".  For example, encrypting to
#        user@example.com will check ldap://keys.example.com.
#
# keyserver = locate a key using whatever keyserver is defined using
#             the keyserver option.
#
# You may also list arbitrary keyservers here by URL.
#
# Try CERT, then PKA, then LDAP, then hkp://subkeys.net:
#auto-key-locate cert pka ldap hkp://subkeys.pgp.net

# default preferences
personal-digest-preferences SHA256
cert-digest-algo SHA256
default-preference-list SHA512 SHA384 SHA256 SHA224 AES256 AES192 AES CAST5 ZLIB BZIP2 ZIP Uncompressed
#+END_SRC

Save and exit.

*** If you have an existing key
#+BEGIN_SRC: bash
gpg --import ~/public_key.txt
gpg --allow-secret-key-import --import ~/private_key.txt
shred -zu ~/private_key.txt
#+END_SRC

Now check the digest preferences, replacing /keyID/ with your GPG key ID.  This applies especially if you have a key which was generated some time ago.

#+BEGIN_SRC: bash
export MYGPGKEYID=keyID
gpg --edit-key $MYGPGKEYID
setpref SHA512 SHA384 SHA256 SHA224 AES256 AES192 AES CAST5 ZLIB BZIP2 ZIP Uncompressed
save
gpg --send-keys $MYGPGKEYID
#+END_SRC
*** To create a new key
Generate a key with the following command:

#+BEGIN_SRC: bash
gpg --gen-key
#+END_SRC

You can find your GPG key ID by entering:

#+BEGIN_SRC: bash
gpg --list-keys
#+END_SRC

The key ID is the second part of the string of numbers and letters.  So for example in:

#+BEGIN_SRC: bash
pub   4096R/EA982E38 2012-05-20
#+END_SRC

the key ID is EA982E38.  Now send your public key to a server so that others can find it.

#+BEGIN_SRC: bash
gpg --send-keys $MYGPGKEYID
#+END_SRC
*** root settings
If you later create an encrypted mailing list then the root user will also need to have good GPG settings so that it can generate key pairs for the list.  The easiest way to ensure this is to do the following, replacing /myusername/ with your username:

#+BEGIN_SRC: bash
su
cp -r /home/myusername/.gnupg ~/
chown -R root:root ~/.gnupg
#+END_SRC

** Protect processes
Because the BBB has limited RAM some processes may occasionally be automatically killed if physical memory availability is getting too low.  The way in which processes are chosen to be sacrificed is not particularly intelligent, and so can result in vital systems being stopped.  To try to prevent that from ever happening the following script can be used, which should ensure that at a minimum ssh, email and mysql keep running.

#+BEGIN_SRC: bash
emacs /usr/bin/protectprocesses
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
#!/bin/bash

declare -a protect=('/usr/sbin/sshd' '/usr/sbin/mysqld --basedir=/usr' '/bin/sh /usr/bin/mysqld_safe' '/usr/sbin/exim4')

for p in "${protect[@]}"
do
  OOM_PROC_ID=$(ps aux | grep '$p' | grep -v grep | head -n 1 | awk -F ' ' '{print $2}')
  if [ ! -z "$OOM_PROC_ID" ]; then
    echo -1000 >/proc/$OOM_PROC_ID/oom_score_adj
    echo -17 >/proc/$OOM_PROC_ID/oom_adj
  fi
done
#+END_SRC

Save and exit, then edit the cron jobs:

#+BEGIN_SRC: bash
emacs /etc/crontab
#+END_SRC

And add the line:

#+BEGIN_SRC: bash
*/1            * *   *   *   root /usr/bin/timeout 30 /usr/bin/protectprocesses
#+END_SRC

Then save and exit and restart cron.

#+BEGIN_SRC: bash
chmod +x /usr/bin/protectprocesses
service cron restart
#+END_SRC

Here cron is used so that if we stop one of the relevant processes and then restart it then its oom priority will be reassigned again
.
** Setting up a web site

#+BEGIN_VERSE
/It's important to have the geek community as a whole think about its responsibility and what it can do. We need various alternative voices pushing back on conventional government sometimes./

-- Tim Berners-Lee
#+END_VERSE

Edit the apache configuration so that it doesn't run out of memory if there are a lot of connections.

#+BEGIN_SRC: bash
su
emacs /etc/apache2/apache2.conf
#+END_SRC

Search for MaxClients and replace the value with 6. As an example the settings should look something like this:

#+BEGIN_SRC: bash
Timeout 30
KeepAlive On
MaxKeepAliveRequests 5
KeepAliveTimeout 1

<IfModule mpm_prefork_module>
    StartServers 5
    MinSpareServers 5
    MaxSpareServers 10
    MaxClients 10
    MaxRequestsPerChild 10
</IfModule>

<IfModule mpm_worker_module>
    StartServers 2
    MinSpareThreads 5
    MaxSpareThreads 5
    ThreadLimit 6
    ThreadsPerChild 10
    MaxClients 10
    MaxRequestsPerChild 10
</IfModule>

<IfModule mpm_event_module>
    StartServers 2
    MinSpareThreads 5
    MaxSpareThreads 5
    ThreadLimit 6
    ThreadsPerChild 10
    MaxClients 10
    MaxRequestsPerChild 10
</IfModule>
#+END_SRC

Also append the following:

#+BEGIN_SRC: bash
ServerSignature Off
ServerTokens Prod
#+END_SRC

Then save and exit.  Install some extra security.

#+BEGIN_SRC: bash
apt-get install libapache2-modsecurity
apt-get install libapache2-mod-evasive
#+END_SRC

In the examples below replace /mydomainname.com/ with your own domain name.

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
mkdir /var/www/$HOSTNAME
mkdir /var/www/$HOSTNAME/htdocs
emacs /etc/apache2/sites-available/$HOSTNAME
#+END_SRC

The Apache configuration for the site should look something like the following.  Replace /mydonainname.com/ with the site domain name.

#+BEGIN_SRC: bash
<VirtualHost *:80>
	ServerAdmin myusername@mydomainname.com
	ServerName mydomainname.com

	DocumentRoot /var/www/mydomainname.com/htdocs
	<Directory />
		Options FollowSymLinks
		AllowOverride All
	</Directory>
	<Directory /var/www/mydomainname.com/htdocs/>
		Options All
		AllowOverride All
		Order allow,deny
		allow from all
        LimitRequestBody 128000
	</Directory>

	# Don't serve .php~ or .php# files created by emacs
	<Files ~ "(^#.*#|~|\.sw[op])$">
		Order allow,deny
		Deny from all
	</Files>

	<IfModule headers_module>
		Header set X-Content-Type-Options nosniff
		Header set Cache-Control "max-age=0, no-cache, no-store, must-revalidate, private"
		Header set Pragma no-cache
	</IfModule>

	<Files .htaccess>
	  deny from all
	</Files>

	ScriptAlias /cgi-bin/ /usr/lib/cgi-bin/
	<Directory "/usr/lib/cgi-bin">
		AllowOverride All
		Options +ExecCGI -MultiViews +SymLinksIfOwnerMatch
		Order allow,deny
		Allow from all
        LimitRequestBody 128000
	</Directory>

	ErrorLog ${APACHE_LOG_DIR}/error.log

	# Possible values include: debug, info, notice, warn, error, crit,
	# alert, emerg.
	LogLevel error

	CustomLog ${APACHE_LOG_DIR}/access.log combined
</VirtualHost>

<IfModule mod_ssl.c>
<VirtualHost *:443>
	ServerAdmin myusername@mydomainname.com
	ServerName mydomainname.com

	DocumentRoot /var/www/mydomainname.com/htdocs
	<Directory />
		Options FollowSymLinks
		AllowOverride All
	</Directory>
	<Directory /var/www/mydomainname.com/htdocs/>
		Options All
		AllowOverride All
		Order allow,deny
		allow from all
        LimitRequestBody 128000
	</Directory>

	# Don't serve .php~ or .php# files created by emacs
	<Files ~ "(^#.*#|~|\.sw[op])$">
		Order allow,deny
		Deny from all
	</Files>

	<IfModule headers_module>
		Header set X-Content-Type-Options nosniff
		Header set Cache-Control "max-age=0, no-cache, no-store, must-revalidate, private"
		Header set Pragma no-cache
	</IfModule>

	<Files .htaccess>
	  deny from all
	</Files>

	ScriptAlias /cgi-bin/ /usr/lib/cgi-bin/
	<Directory "/usr/lib/cgi-bin">
		AllowOverride All
		Options +ExecCGI -MultiViews +SymLinksIfOwnerMatch
		Order allow,deny
		Allow from all
        LimitRequestBody 128000
	</Directory>

	ErrorLog ${APACHE_LOG_DIR}/error.log

	# Possible values include: debug, info, notice, warn, error, crit,
	# alert, emerg.
	LogLevel error

	CustomLog ${APACHE_LOG_DIR}/ssl_access.log combined

	#   SSL Engine Switch:
	#   Enable/Disable SSL for this virtual host.
	SSLEngine on

	#   A self-signed certificate
	SSLCertificateFile    /etc/ssl/certs/mydomainname.com.crt
	SSLCertificateKeyFile /etc/ssl/private/mydomainname.com.key

    # Options based on bettercrypto.org
    SSLProtocol All -SSLv2 -SSLv3
    SSLHonorCipherOrder On
    SSLCompression off
    SSLCipherSuite EDH+CAMELLIA:EDH+aRSA:EECDH+aRSA+AESGCM:EECDH+aRSA+SHA384:EECDH+aRSA+SHA256:EECDH:+CAMELLIA256:+AES256:+CAMELLIA128:+AES128:+SSLv3:!aNULL:!eNULL:!LOW:!3DES:!MD5:!EXP:!PSK:!SRP:!DSS:!RC4:!SEED:!ECDSA:CAMELLIA256-SHA:AES256-SHA:CAMELLIA128-SHA:AES128-SHA

    # Add six earth month HSTS header for all users ...
    Header add Strict-Transport-Security "max-age=15768000"
    # If you want to protect all subdomains , use the following header
    # ALL subdomains HAVE TO support https if you use this !
    # Strict-Transport-Security: max-age=15768000 ; includeSubDomains

	#   SSL Engine Options:
	#   Set various options for the SSL engine.
	#   o FakeBasicAuth:
	#     Translate the client X.509 into a Basic Authorisation.  This means that
	#     the standard Auth/DBMAuth methods can be used for access control.  The
	#     user name is the `one line' version of the client's X.509 certificate.
	#     Note that no password is obtained from the user. Every entry in the user
	#     file needs this password: `xxj31ZMTZzkVA'.
	#   o ExportCertData:
	#     This exports two additional environment variables: SSL_CLIENT_CERT and
	#     SSL_SERVER_CERT. These contain the PEM-encoded certificates of the
	#     server (always existing) and the client (only existing when client
	#     authentication is used). This can be used to import the certificates
	#     into CGI scripts.
	#   o StdEnvVars:
	#     This exports the standard SSL/TLS related `SSL_*' environment variables.
	#     Per default this exportation is switched off for performance reasons,
	#     because the extraction step is an expensive operation and is usually
	#     useless for serving static content. So one usually enables the
	#     exportation for CGI and SSI requests only.
	#   o StrictRequire:
	#     This denies access when "SSLRequireSSL" or "SSLRequire" applied even
	#     under a "Satisfy any" situation, i.e. when it applies access is denied
	#     and no other module can change it.
	#   o OptRenegotiate:
	#     This enables optimized SSL connection renegotiation handling when SSL
	#     directives are used in per-directory context.
	#SSLOptions +FakeBasicAuth +ExportCertData +StrictRequire
	<FilesMatch "\.(cgi|shtml|phtml|php)$">
		SSLOptions +StdEnvVars
	</FilesMatch>
	<Directory /usr/lib/cgi-bin>
		SSLOptions +StdEnvVars
	</Directory>

	#   SSL Protocol Adjustments:
	#   The safe and default but still SSL/TLS standard compliant shutdown
	#   approach is that mod_ssl sends the close notify alert but doesn't wait for
	#   the close notify alert from client. When you need a different shutdown
	#   approach you can use one of the following variables:
	#   o ssl-unclean-shutdown:
	#     This forces an unclean shutdown when the connection is closed, i.e. no
	#     SSL close notify alert is send or allowed to received.  This violates
	#     the SSL/TLS standard but is needed for some brain-dead browsers. Use
	#     this when you receive I/O errors because of the standard approach where
	#     mod_ssl sends the close notify alert.
	#   o ssl-accurate-shutdown:
	#     This forces an accurate shutdown when the connection is closed, i.e. a
	#     SSL close notify alert is send and mod_ssl waits for the close notify
	#     alert of the client. This is 100% SSL/TLS standard compliant, but in
	#     practice often causes hanging connections with brain-dead browsers. Use
	#     this only for browsers where you know that their SSL implementation
	#     works correctly.
	#   Notice: Most problems of broken clients are also related to the HTTP
	#   keep-alive facility, so you usually additionally want to disable
	#   keep-alive for those clients, too. Use variable "nokeepalive" for this.
	#   Similarly, one has to force some clients to use HTTP/1.0 to workaround
	#   their broken HTTP/1.1 implementation. Use variables "downgrade-1.0" and
	#   "force-response-1.0" for this.
	BrowserMatch "MSIE [2-6]" \
		nokeepalive ssl-unclean-shutdown \
		downgrade-1.0 force-response-1.0
	# MSIE 7 and newer should be able to use keepalive
	BrowserMatch "MSIE [17-9]" ssl-unclean-shutdown

</VirtualHost>
</IfModule>
#+END_SRC

Then to enable the site:

#+BEGIN_SRC: bash
a2ensite
a2dissite default
a2dissite default-ssl
a2enmod rewrite
a2enmod headers
#+END_SRC

Ensure that "NameVirtualHost *:443" is added to /etc/apache2/ports.conf. It should look something like the following:

#+BEGIN_SRC: bash
NameVirtualHost *:80
Listen 80

<IfModule mod_ssl.c>
	NameVirtualHost *:443
    Listen 443
</IfModule>

<IfModule mod_gnutls.c>
    NameVirtualHost *:443
    Listen 443
</IfModule>
#+END_SRC

Create a self-signed certificate. The passphrase isn't important and will be removed, so make it easy (such as "password").

#+BEGIN_SRC: bash
emacs /usr/bin/makecert
#+END_SRC

Enter the following:

#+BEGIN_SRC: bash
#!/bin/bash

HOSTNAME=$1

openssl genrsa -des3 -out $HOSTNAME.key 1024
openssl req -new -x509 -nodes -sha1 -days 3650 -key $HOSTNAME.key -out $HOSTNAME.crt
openssl rsa -in $HOSTNAME.key -out $HOSTNAME.new.key
cp $HOSTNAME.new.key $HOSTNAME.key
rm $HOSTNAME.new.key
cp $HOSTNAME.key /etc/ssl/private
chmod 400 /etc/ssl/private/$HOSTNAME.key
cp $HOSTNAME.crt /etc/ssl/certs
shred -zu $HOSTNAME.key $HOSTNAME.crt
a2enmod ssl
service apache2 restart
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod +x /usr/bin/makecert
makecert mydomainname.com
#+END_SRC

Enter some trivial password for the key file, such as "password".  The password will be removed as part of the /makecert/ script which you just created.  Note that leaving a password on the key file would mean that after a power cycle the Apache server will not be able to boot properly (it would wait indefinitely for a password to be manually entered) and would look as if it had crashed.

If all has gone well then there should be no warnings or errors after you run the service restart command.  After that you should enable ports 80 (HTTP) and 443 (HTTPS) on your internet router/firewall, such that they are redirected to the BBB.

Also limit the amount of memory which any php scripts can use.

#+BEGIN_SRC: bash
emacs /etc/php5/apache2/php.ini
#+END_SRC

Set the following:

#+BEGIN_SRC: bash
memory_limit = 32M
#+END_SRC

Save and exit.  Also edit */etc/php5/cli/php.ini* and set /memory_limit/ to the same value.  This should prevent any rogue scripts from crashing the system.

** Accessing your Email

#+BEGIN_VERSE
/The emails showed that Google...was among several other military/defense contractors vying for a piece of DAC’s $10.9-million surveillance contracting action./

-- Article on the "Google-Military-Surveillance Complex" by Yasha Levine
#+END_VERSE

*** Mutt email client

#+BEGIN_SRC: bash
apt-get install mutt-patched lynx abook
exit
mkdir ~/.mutt
echo "text/html; lynx -dump -width=78 -nolist %s | sed ‘s/^ //’; copiousoutput; needsterminal; nametemplate=%s.html" > ~/.mutt/mailcap
su
emacs /etc/Muttrc
#+END_SRC

Append the following:

#+BEGIN_SRC: bash
set mbox_type=Maildir
set folder="~/Maildir"
set mask="!^\\.[^.]"
set mbox="~/Maildir"
set record="+Sent"
set postponed="+Drafts"
set trash="+Trash"
set spoolfile="~/Maildir"
auto_view text/x-vcard text/html text/enriched
set editor="emacs"
set header_cache="+.cache"

macro index S "<tag-prefix><save-message>=.learn-spam<enter>" "move to learn-spam"
macro pager S "<save-message>=.learn-spam<enter>" "move to learn-spam"
macro index H "<tag-prefix><copy-message>=.learn-ham<enter>" "copy to learn-ham"
macro pager H "<copy-message>=.learn-ham<enter>" "copy to learn-ham"

# set up the sidebar
set sidebar_width=12
set sidebar_visible=yes
set sidebar_delim='|'
set sidebar_sort=yes

set rfc2047_parameters

# Show inbox and sent items
mailboxes = =Sent

# Alter these colours as needed for maximum bling
color sidebar_new yellow default
color normal white default
color hdrdefault brightcyan default
color signature green default
color attachment brightyellow default
color quoted green default
color quoted1 white default
color tilde blue default

# ctrl-n, ctrl-p to select next, prev folder
# ctrl-o to open selected folder
bind index \Cp sidebar-prev
bind index \Cn sidebar-next
bind index \Co sidebar-open
bind pager \Cp sidebar-prev
bind pager \Cn sidebar-next
bind pager \Co sidebar-open

# ctrl-b toggles sidebar visibility
macro index,pager \Cb '<enter-command>toggle sidebar_visible<enter><redraw-screen>' "toggle sidebar"

# esc-m Mark new messages as read
macro index <esc>m "T~N<enter>;WNT~O<enter>;WO\CT~T<enter>" "mark all messages read"


# Collapsing threads
macro index [ "<collapse-thread>" "collapse/uncollapse thread"
macro index ] "<collapse-all>"    "collapse/uncollapse all threads"

# threads containing new messages
uncolor index "~(~N)"
  color index brightblue default "~(~N)"

# new messages themselves
uncolor index "~N"
  color index brightyellow default "~N"


# GPG/PGP integration

# this set the number of seconds to keep in memory the passphrase used to encrypt/sign
set pgp_timeout=60

# automatically sign and encrypt
set pgp_autosign         # autosign all outgoing mails
set pgp_replyencrypt     # autocrypt replies to crypted
set pgp_replysign        # autosign replies to signed
set pgp_auto_decode=yes  # decode attachments
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/mail/spamassassin/local.cf
#+END_SRC

Uncomment *use_bayes*, *bayes_auto_learn*

Save and exit, then run:

#+BEGIN_SRC: bash
service spamassassin restart
#+END_SRC

Now to add an address book:

#+BEGIN_SRC: bash
emacs ~/.muttrc
#+END_SRC

Append the following:

#+BEGIN_SRC: bash
set alias_file=~/.mutt-alias
source ~/.mutt-alias
set query_command= "abook --mutt-query '%s'"
macro index,pager A "<pipe-message>abook --add-email-quiet<return>" "add the sender address to abook"
#+END_SRC

Then save and exit.

#+BEGIN_SRC: bash
touch ~/.mutt-alias
#+END_SRC

Finally you can then type *mutt* to get access to your email.  Hence as a fallback, or if you prefer as the primary way of accessing email, you can ssh into the BBB and use the mutt command line email client.  Ssh clients are available for all operating systems, and also you should be reasonably protected from passive surveillance between wherever you are and the BBB (although not between the BBB and the wider internet), which can be useful if you are for example using an Android tablet from a cafe or railway station.

To use the address book system open an email and then to add the sender to the address list press the A key.  It will ask you for an alias which may be used the next time you want to send a mail.  Alternatively you may just edit the *~/.mutt-alias* file directly to add email addresses.

Some useful keys to know are:

| ESC /  | Search for text within message contents                 |
| "/"    | Search for text within headers                          |
| *      | Move to the last message                                |
| TAB    | Move to the next unread message                         |
| d      | Delete a message                                        |
| u      | Undelete a mail which is pending deletion               |
| $      | Delete all messages selected and check for new messages |
| a      | Add to the address book                                 |
| m      | Send a new mail                                         |
| ESC-m  | Mark all messages as having been read                   |
| S      | Mark a message as spam                                  |
| H      | Mark a message as ham                                   |
| CTRL-b | Toggle side bar on/off                                  |
| CTRL-n | Next mailbox (on side bar)                              |
| CTRL-p | Previous mailbox (on side bar)                          |
| CTRL-o | Open mailbox (on side bar)                              |
| ]      | Expand or collapse all threads                          |
| [      | Expand of collapse the current thread                   |
| CTRL-k | Import a PGP/GPG public key                             |

One of the most common things which you might wish to do is to send an email.  To do this first press /m/ to create a new message.  Enter the address to send to and the subject, then after a few seconds the Emacs editor will appear with a blank document.  Type your email then press /CTRL-x CTRL-s/ to save it and /CTRL-x CTRL-c/ to exit.  You will then see a summary of the email to be sent out.  Press /y/ to send it and then enter your GPG key passphrase (the one you gave when creating a PGP/GPG key).  The purpose of that is to add a signature which is a strong proof that the email was written by you and not by someone else.

*** K9 Android client

#+BEGIN_VERSE
/The surveillance state is robust. It is robust politically, legally, and technically./

-- Bruce Schneier
#+END_VERSE

**** Incoming server settings
 * Select settings/account settings
 * Select Fetching mail/incoming server
 * Enter your username and password
 * IMAP server should be your domain name
 * Security: SSL/TLS (always)
 * Authentication: Plain
 * Port: 993
**** Outgoing (SMTP) server settings
 * Select settings/account settings
 * Select Sending mail/outgoing server
 * Set SMTP server to your domain name
 * Set Security to SSL/TLS (always)
 * Set port to 465
 * Set authentication to PLAIN
 * Enter your username and password
 * Accept the SSL certificate
**** Folders
To view any new folders which you may have created using the /mailinglistrule/ script from your inbox press the *K9 icon* at the top left to access folders, then press the *menu button* and select *refresh folder list*.

If your folder still doesn't show up then press the *menu button*, select *show folders* and select *all folders*.
*** Webmail

#+BEGIN_VERSE
/Most of the information extracted is "content", such as recordings of phone calls or the substance of email messages./

-- From a 2013 Guardian article on GCHQ/NSA bulk internet data interception.
#+END_VERSE

For maximum speed and efficiency the recommended email client is Mutt, accessed via ssh, but non-technical people who aren't using an Android app are unlikely to want to use email in that manner.  So it's a good idea to also have a webmail system installed, both for accessibility and as a fallback should ssh not be available due to port blocking.

Install dependencies.

#+BEGIN_SRC: bash
apt-get install libapache2-mod-authz-unixgroup
#+END_SRC

Create a mysql database, specifying a password which should be a long random string generated with a password manager such as KeepassX.

#+BEGIN_SRC: bash
mysql -u root -p
create database roundcubemail;
CREATE USER 'roundcube'@'localhost' IDENTIFIED BY 'roundcubepassword';
GRANT ALL PRIVILEGES ON roundcubemail.* TO 'roundcube'@'localhost';
quit
#+END_SRC

Download roundcube.

#+BEGIN_SRC: bash
cd /tmp
wget http://freedombone.uk.to/roundcubemail.tar.gz
#+END_SRC

Verify it.

#+BEGIN_SRC: bash
sha256sum roundcubemail.tar.gz
e8a311b22a8e1f70abb72ed9551cc9233cf6c5221f1eebf1ae64974117e3148b roundcubemail.tar.gz
#+END_SRC

Extract the files.

#+BEGIN_SRC: bash
tar -xzvf roundcubemail.tar.gz
export HOSTNAME=mydomainname.com
cp -r roundcubemail-* /var/www/$HOSTNAME/htdocs/mail
chown -R www-data:www-data /var/www/$HOSTNAME/htdocs/mail/temp
chown -R www-data:www-data /var/www/$HOSTNAME/htdocs/mail/logs
rm /var/www/$HOSTNAME/htdocs/mail/.htaccess
#+END_SRC

Edit your web site configuration.

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/$HOSTNAME
#+END_SRC

Within the 80 VirtualHost section add the following:

#+BEGIN_SRC: bash
  <Directory /var/www/mydomainname.com/htdocs/mail>
    deny from all
  </Directory>
#+END_SRC

Within the 443 VirtualHost section add the following:

#+BEGIN_SRC: bash
  <Directory /var/www/mydomainname.com/htdocs/mail>
    Options Indexes FollowSymLinks MultiViews
    AllowOverride All
    Order allow,deny
    allow from all
  </Directory>
#+END_SRC

Save and exit, then restart Apache.

#+BEGIN_SRC: bash
service apache2 restart
#+END_SRC

Now with a browser visit https://mydomainname.com/mail/installer. Scroll down and click "next".  Give your webmail site a product name.

The *spellcheck_engine* option being limited to Google is slightly concerning in terms of privacy and security, but seems not to be implemented.

Change the *database password* to the password you gave when creating the MySql database above.

Set *smtp_port* to 465.

Click *create config*

Click download to download the file.

The config file which you downloaded should contain the following:

#+BEGIN_SRC: bash
$config['default_host'] = 'localhost';
$config['smtp_port'] = 465;
$config['username_domain'] = '';
#+END_SRC

In a terminal on your local machine (not logged into the BBB):

#+BEGIN_SRC: bash
cd ~/Downloads
scp config.inc.php myusername@mydomainname.com:/home/myusername
#+END_SRC

Then in a terminal ssh'd into the BBB:

#+BEGIN_SRC: bash
mv /home/myusername/config.inc.php /var/www/$HOSTNAME/htdocs/mail/config
chmod 755 /var/www/$HOSTNAME/htdocs/mail/config/config.inc.php
#+END_SRC

Click *continue*.

Click *initialize database*.

Under *Test SMTP config* you can use a [[mailinator.com]] address to check that mail can be sent.

Now we can delete the installer.

#+BEGIN_SRC: bash
rm -rf /var/www/$HOSTNAME/htdocs/mail/installer
#+END_SRC

Now with a browser navigate to https://mydomainname.com/mail and log in.

You'll notice that you may not be able to see any mailing list folders which you may have created earlier using the /mailinglistrule/ script. To make folders visible click on the cog-like settings icon at the bottom left of the screen then select *manage folders*.  You will then be able to select which folders you wish to become visible.  Make sure that the *Sent*, *spam* and *ham* folders are selected.

Click on the *Mail* icon to go back to your main mail screen then click on the *Settings* icon at the top right of the screen and select *special folders*. Set *Junk* to *spam* then click the save button. Also select *identities* and make sure that your email address is correct.

*** Thunderbird

#+BEGIN_VERSE
/Towards the end of 2012, we heard from the National Technical Assistance Centre (NTAC), a division of GCHQ and a liaison with the Home Office, [that] they wanted the keys to decrypt the customer data./

-- Brian Spector, on the shutting down of the PrivateSky encrypted email service
#+END_VERSE

Another common way in which you may want to access email is via Thunderbird.  This may be especially useful if you're trying to convert former Windows users who may previously have been using some version of Outlook.

The following instructions should be carried out on the client machines (laptop, etc), not on the BBB itself.

**** Initial setup

Install *Thunderbird* and *Enigmail*.  How you do this just depends upon your distro and software manager or "app store".

Open Thinderbird

Select "*Skip this and use existing email*"

Enter your name, email address (myusername@mydomainname.com) and the password for your user (the one from [[Add a user]]).

You'll get a message saying "/Thunderbird failed to find the settings/"

The settings should be as follows, substituting /mydomainname.com/ for your domain name and /myusername/ for the username given previously in [[Add a user]].

  * Incoming: IMAP, mydomainname.com, 993, SSL/TLS, Normal Password
  * Outgoing: SMTP, mydomainname.com, 465, SSL/TLS, Normal Password
  * Username: myusername

Click *Done*.

Click *Get Certificate* and make sure "*permanently store this exception*" is selected", then click *Store Security Exception*.

From OpenPGP setup select "*Yes, I would like the wizard to get me started*".  If the wizard doesn't start automatically then "setup wizard" can be selected from OpenPGP on the menu bar.

Select "*Yes, I want to sign all of my email*"

Select "*No, I will create per-recipient rules*"

Select "*yes*" to change default settings.

**** If you have existing GPG key

Export your GPG public and private keys.

#+BEGIN_SRC: bash
gpg --output ~/public_key.txt --armor --export KEY_ID
gpg --output ~/private_key.txt --armor --export-secret-key KEY_ID
#+END_SRC

Select "*I have existing public and private keys*".

Select your public and private GPG exported key files.

Select the account which you want to use and click *Next*, *Next* and *Finish*.

Remove your exported key files.

#+BEGIN_SRC: bash
shred -zu ~/public_key.txt
shred -zu ~/private_key.txt
#+END_SRC

**** If you don't have any existing GPG or PGP key
Select "*I want to create a new key pair*"

Enter a passphrase and click *Next* a couple of times.

Click *Generate Certificate* to generate a revocation certificate.

Enter the passphrase which you gave previously.

Click *Finish*

From the menu select *OpenPGP* and then *Key Management*.  Make sure that *Display all keys* is selected and then select your key. Select *Keyserver* on the menu and then *Upload Public Keys*.  This will upload your public key to a key server so that others can find it.

Select *File* from the menu then *Export keys to file*.  Click on *Export Secret keys* and select a location to save them to.  It's a good idea to save them to a USB stick which can then be removed from the computer and carried around on a keyring together with your physical keys.  If you need to set up GPG or Thunderbird/Enigmail on others then this file will be used to import your keys.

**** Using for the first time

Click on the Thunderbird menu, which looks like three horizontal bars on the right hand side.

Hover over *preferences* and then *Account settings*.

Select *Synchronization & Storage*.

Make sure that *Keep messages for this account on this computer* is unticked, then click *Ok*.

Click on *Inbox*.  Depending upon how much email you have it may take a while to import the subject lines.

Note that when sending an email for the first time you will also need to accept the SSL certificate.

Get into the habit of using email encryption and encourage others to do so.  Remember that you may not think that your emails are very interesting but the Surveillance State is highly interested in them and will be actively trying to data mine your private life looking for "suspicious" patterns, regardless of whether you are guilty of any crime or not.

**** Making folders visible
By default you won't be able to see any folders which you may have created earlier using the /mailinglistrule/ script.  To make folders visible select:

*Menu*, hover over *Preferences*, select *Account Settings*, select *Server Settings* then click on the *Advanced* button.

Make sure that "*show only subscribed folders*" is not checked.  Then click the *ok* buttons.  Folders will be re-scanned, which may take some time depending upon how much email you have, but your folders will then appear.
** Create Email folders and rules

#+BEGIN_VERSE
/Yes, the NSA set fire to the Internet but it’s the business models of Google, Facebook, etc, that provide the firewood.  Trusting the companies supplying the firewood to be your fire fighters is naïve at best./

-- Aral Balkan
#+END_VERSE

*** Rules for mailing lists
A common situation with email is that you may be subscribed to various mailing lists and want incoming email from those to be automatically grouped into a separate folder for each list.

We can make a script to make adding mailing list rules easy:

#+BEGIN_SRC: bash
emacs /usr/bin/mailinglistrule
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
#!/bin/bash
MYUSERNAME=$1
MAILINGLIST=$2
SUBJECTTAG=$3
MUTTRC=/home/$MYUSERNAME/.muttrc
PM=/home/$MYUSERNAME/.procmailrc
LISTDIR=/home/$MYUSERNAME/Maildir/$MAILINGLIST
if [ ! -d "$LISTDIR" ]; then
  mkdir -m 700 $LISTDIR
  mkdir -m 700 $LISTDIR/tmp
  mkdir -m 700 $LISTDIR/new
  mkdir -m 700 $LISTDIR/cur
fi
chown -R $MYUSERNAME:$MYUSERNAME $LISTDIR
echo "" >> $PM
echo ":0" >> $PM
echo "  * ^Subject:.*()\[$SUBJECTTAG\]" >> $PM
echo "$LISTDIR/new" >> $PM
chown $MYUSERNAME:$MYUSERNAME $PM
if [ ! -f "$MUTTRC" ]; then
  cp /etc/Muttrc $MUTTRC
  chown $MYUSERNAME:$MYUSERNAME $MUTTRC
fi
PROCMAILLOG=/home/$MYUSERNAME/log
if [ ! -d $PROCMAILLOG ]; then
  mkdir $PROCMAILLOG
  chown -R $MYUSERNAME:$MYUSERNAME $PROCMAILLOG
fi
#+END_SRC

Save and exit, then make the script executable.

#+BEGIN_SRC: bash
chmod +x /usr/bin/mailinglistrule
#+END_SRC

Now we can add a new mailing list rule with the following, where /myusername/ is your username, /mailinglistname/ is the name of the mailing list (with no spaces) and /subjecttag/ is the tag which usually appears within square brackets in the subject line of emails from the list.

#+BEGIN_SRC: bash
mailinglistrule [myusername] [mailinglistname] [subjecttag]
#+END_SRC

Repeat this command for as many mailing lists as you need.  Then edit your local Mutt configuration.

#+BEGIN_SRC: bash
emacs /home/myusername/.muttrc
#+END_SRC

Search for the *mailboxes* variable and add entries for the mailing lists you just created.  For example:

#+BEGIN_SRC: bash
mailboxes = =Sent =mailinglistname
#+END_SRC

Then save and exit.

*** Rules for specific email addresses

You can also make a script which will allow you to move mail from specific email addresses to a folder.

#+BEGIN_SRC: bash
emacs /usr/bin/emailrule
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
#!/bin/bash
MYUSERNAME=$1
EMAILADDRESS=$2
MAILINGLIST=$3
MUTTRC=/home/$MYUSERNAME/.muttrc
PM=/home/$MYUSERNAME/.procmailrc
LISTDIR=/home/$MYUSERNAME/Maildir/$MAILINGLIST
if [ ! -d "$LISTDIR" ]; then
  mkdir -m 700 $LISTDIR
  mkdir -m 700 $LISTDIR/tmp
  mkdir -m 700 $LISTDIR/new
  mkdir -m 700 $LISTDIR/cur
fi
chown -R $MYUSERNAME:$MYUSERNAME $LISTDIR
echo "" >> $PM
echo ":0" >> $PM
echo "  * ^From: $EMAILADDRESS" >> $PM
echo "$LISTDIR/new" >> $PM
chown $MYUSERNAME:$MYUSERNAME $PM
if [ ! -f "$MUTTRC" ]; then
  cp /etc/Muttrc $MUTTRC
  chown $MYUSERNAME:$MYUSERNAME $MUTTRC
fi
PROCMAILLOG=/home/$MYUSERNAME/log
if [ ! -d $PROCMAILLOG ]; then
  mkdir $PROCMAILLOG
  chown -R $MYUSERNAME:$MYUSERNAME $PROCMAILLOG
fi
#+END_SRC

Save and exit, then make the script executable.

#+BEGIN_SRC: bash
chmod +x /usr/bin/emailrule
#+END_SRC

Then to add a particular email address to a folder run the command:

#+BEGIN_SRC: bash
emailrule [myusername] [emailaddress] [foldername]
#+END_SRC

If you want any mail from the given email address to be deleted then set the /foldername/  to /Trash/.

To ensure that the folder appears within Mutt.

#+BEGIN_SRC: bash
emacs /home/myusername/.muttrc
#+END_SRC

Search for the *mailboxes* variable and add entries for the mailing lists you just created.  For example:

#+BEGIN_SRC: bash
mailboxes = =Sent =foldername
#+END_SRC

Then save and exit.

** Install a Blog

#+BEGIN_VERSE
/When society gives censors wide and vague powers they never confine themselves to deserving targets. They are not snipers, but machine-gunners. Allow them to fire at will, and they will hit anything that moves./

-- Nick Cohen
#+END_VERSE

Wordpress is the most popular blogging platform, but in practice I found it to be high maintenance with frequent security updates and breakages. More practical for a home server is Flatpress.  Flatpress doesn't use a MySql database, just text files, and so is easy to relocate or reinstall.

See the [[Setting up a web site]] section of this document for details of how to configure the web server for your blog's domain.

Download flatpress.

#+BEGIN_SRC: bash
cd /tmp
wget http://freedombone.uk.to/flatpress.tar.gz
#+END_SRC

Verify the download:

#+BEGIN_SRC: bash
sha256sum flatpress.tar.gz
6312a49aab5aabd6371518dcaf081f489dff04d001bc34b4fe3f2a81170bbd4e flatpress.tar.gz
#+END_SRC

Extract and install it.

#+BEGIN_SRC: bash
tar -xzvf flatpress.tar.gz
cd flatpress-*
cp -r * /var/www/$HOSTNAME/htdocs
cd ..
rm -rf flatpress-*
rm -f flatpress.tar.gz
#+END_SRC

Now visit your blog and follow the setup instructions, which are quite minimal.  Various themes and addons are available from the Flatpress web site, http://www.flatpress.org

** Install an IRC server

#+BEGIN_VERSE
/Give me the liberty to know, to utter, and to argue freely according to conscience, above all liberties./

-- John Milton
#+END_VERSE

*** Base install

IRC is not an especially secure system.  For instance, even with the best encryption it's easily possible to imagine IRC-specific cribs which could be used by cryptanalytic systems.  However, we'll try to implement it in a manner which will at least give the surveillance aparatus something to ponder over.

First install some dependencies.

#+BEGIN_SRC: bash
apt-get update
apt-get install build-essential openssl libssl-dev debhelper dpatch docbook-to-man flex bison libpcre3-dev
#+END_SRC

Then get the source code for ircd-hybrid.

#+BEGIN_SRC: bash
cd /tmp
mkdir hybrid
cd hybrid
apt-get source ircd-hybrid
#+END_SRC

Modify the source code to include SSL security.

#+BEGIN_SRC: bash
emacs ircd-hybrid-*/debian/rules
#+END_SRC

Beneath MAXCLIENTS add the line:

#+BEGIN_SRC: bash
USE_OPENSSL = 1
#+END_SRC

Then save and exit.  Now we can build the debian package for ircd-hybrid and install it.

#+BEGIN_SRC: bash
cd ircd-hybrid-*
dpkg-buildpackage -rfakeroot -uc -b
cd ..
dpkg -i ircd-hybrid_*.deb
#+END_SRC

Customise the configuration to your system, giving it a name and description.  In this example 192.168.1.60 is the static IP address on the BBB on the local network, so change that if necessary.

#+BEGIN_SRC: bash
emacs /etc/ircd-hybrid/ircd.conf
#+END_SRC

Set *name* to the name of your server, and set a description.

Set a *network_name* and *network_desc*.  The network name should not contain any spaces.

Set max_clients to 20, or however many you expect that you'll typically need.

Within the admin section set your *name* and *email*.

Within the *listen* section set host to your fixed IP address (in the earlier
sections it was 192.168.1.60).

Within the *auth* section set user = "*@192.168.1.60" - or whatever the fixed IP address of the BBB is on your network.

Uncomment the first *connect* section and set the *name* to your domain name, the *host* to 192.168.1.60 and the send/accept passwords to a password which you use to log into the IRC server.  Also set the *port* to 6670.

Save and exit, then restart the IRC server.  Open port 6670 on your internet router and forward it to the BBB.

Ensure that the configuration is only readable by the root user.

#+BEGIN_SRC: bash
chmod 600 /etc/ircd-hybrid/ircd.conf
#+END_SRC

*** Channel management

To to install channel management tools.

#+BEGIN_SRC: bash
cd /tmp
wget http://freedombone.uk.to/hybserv_1.9.4-1_armhf.deb
#+END_SRC

Verify it.

#+BEGIN_SRC: bash
sha256sum hybserv_1.9.4-1_armhf.deb
41bf4eb6e24c87610a80bc14db1103a57484835510eea7e4ba9709c523318615 hybserv_1.9.4-1_armhf.deb
#+END_SRC

Install it.

#+BEGIN_SRC: bash
dpkg -i hybserv_1.9.4-1_armhf.deb
#+END_SRC

Make a md5 version of the password for the IRC server operator.

#+BEGIN_SRC: bash
/usr/bin/mkpasswd <myoperatorpassword>
#+END_SRC

Edit the ircd-hybrid configuration.

#+BEGIN_SRC: bash
emacs /etc/ircd-hybrid/ircd.conf
#+END_SRC

Enter the md5 password which you previously created within the /operator/ section.  Also change /user/ to:

#+BEGIN_SRC: bash
	user = "*@*";
#+END_SRC

Then save and exit.

#+BEGIN_SRC: bash
emacs /etc/hybserv/hybserv.conf
#+END_SRC

Change #MD5 PASSWORD HERE# to the md5 operator password created earlier, mydomainname.com to your domain name and mysendacceptpassword to the send/accept password specified within /ircd.conf/.

#+BEGIN_SRC: bash
A:mynickname <myemailaddress>
N:irc.mydomainname.com:Hybrid services
O:*@*:#MD5 PASSWORD HERE#:root:segj (comment out other Q: lines)
S:mysendacceptpassword:192.168.1.60:6670 (remove the other two services)
#+END_SRC

Also remove the line *#NOT-EDITED#*, then save and exit.

Now we need to restart the ircd and hybrid server to make things work:

#+BEGIN_SRC: bash
service ircd-hybrid restart
service hybserv start
#+END_SRC

*** Usage with Irssi

On another computer (not the BBB).

#+BEGIN_SRC: bash
sudo apt-get install irssi irssi-plugin-otr irssi-plugin-xmpp
irssi
#+END_SRC

Connect to the IRC and identify yourself as an operator.  Here /mynetwork/ should be the same as *network_name* specified earlier within /ircd.conf/.  The network name is something equivalent to "freenode".

#+BEGIN_SRC: bash
/network add -nick mynick mynetwork

/channel add -auto #mychannel mynetwork channelpassword

/server add -auto -network mynetwork -ssl mydonainname.com 6670 mysendacceptpassword

/connect mydomainname.com

/join #mychannel

/msg -servername chanserv REGISTER #mychannel channelpassword

/msg -servername chanserv set #mychannel mlock +k channelpassword
#+END_SRC

If you edit the irssi config file:

#+BEGIN_SRC: bash
emacs ~/.irssi/config
#+END_SRC

It should look something like this:

#+BEGIN_SRC: bash
  {
    address = "mydomainname.com";
    chatnet = "mynetwork";
    port = "6670";
    password = "mysendacceptpassword";
    use_ssl = "yes";
    ssl_verify = "no";
    autoconnect = "yes";
  },
#+END_SRC

If you're not using a self-signed certificate (self-signed is the default) then you can set *ssl_verify* to "yes".

By default irssi will use UTC time.  An example of setting to some other time zone is as follows:

#+BEGIN_SRC: bash
echo "load perl" >> ~/.irssi/startup
echo "script exec $ENV{'TZ'}='Europe/London';" >> ~/.irssi/startup
#+END_SRC

Also enable /Off The Record/ (OTR) messaging.

#+BEGIN_SRC: bash
echo "load otr" >> ~/.irssi/startup
#+END_SRC

By default Irssi does not look especially attractive.  To improve it's looks:

#+BEGIN_SRC: bash
cd ~/.irssi
wget http://freedombone.uk.to/irssi/xchat.theme
mkdir ~/.irssi/scripts
mkdir ~/.irssi/scripts/autorun
cd ~/.irssi/scripts/autorun
wget http://freedombone.uk.to/irssi/xchatnickcolor.pl
wget http://freedombone.uk.to/irssi/adv_windowlist.pl
#+END_SRC

Verify the files:

#+BEGIN_SRC: bash
sha256sum ~/.irssi/xchat.theme
7a84130ad55aabd0b043a03b013628438e6c7f82a58e15267633bc7eb443e60b

sha256sum ~/.irssi/scripts/autorun/xchatnickcolor.pl
8293e867a22d42ce5a28cd755237509b6f3587fd2b21d7d20af4a832081610ca

sha256sum ~/.irssi/scripts/autorun/adv_windowlist.pl
e4dd8f6d384bf4f2d0ab5ccf06df06e4a69d2647b08d37c8fc6cfd9326688395
#+END_SRC

Then run Irssi and enter the commands:

#+BEGIN_SRC: bash
/set theme xchat
/statusbar window remove act
/set awl
/set awl_block -14
/set awl_display_key $Q%K|$N%n $H$C$S
/set awl_display_key_active $Q%K|$N%n $H%U$C%n$S
/set awl_display_nokey [$N]$H$C$S
/run autorun/adv_windowlist.pl
/set awl_viewer off
/save
#+END_SRC

*** Using irssi with Off The Record messaging (OTR)
Once you're running irssi then you can enable OTR with:

#+BEGIN_SRC: bash
/statusbar window add otr
/otr genkey mynick@network  (for example mynick@irc.freenode.net)
#+END_SRC

Then to see your OTR fingerprint:

#+BEGIN_SRC: bash
/otr info
#+END_SRC

And to trust or distrust someone else's fingerprint.

#+BEGIN_SRC: bash
/otr trust [fingerprint]
/otr distrust [fingerprint]
#+END_SRC

*** Usage with XChat
Within the network list click, *Add* and enter your domain name then click *Edit*.

Select the entry within the servers box, then enter *mydomainname.com/6670* and press *Enter*.

Uncheck *use global user information*.

Enter first and second nicknames and check *auto connect to this network on startup*.

Check *use SSL* and *accept invalid SSL certificate*.

Enter some favourite channels and within *server password* enter /mysendacceptpassword/ which you defined earlier when setting up the server.

Click *close* and then *connect*.
*** Install Irssi as a daemon

It may be useful to run a persistent Irssi session on the BBB.  This will enable you to log in and see any entries which occurred previously so that you don't find yourself in an argument without knowledge of what was said in the last few minutes or hours.  This feature only works for a single user on the BBB - typically the administrator.

First install some prerequisites.

#+BEGIN_SRC: bash
apt-get install irssi irssi-plugin-otr irssi-plugin-xmpp screen
#+END_SRC

Create an initialisation script.

#+BEGIN_SRC: bash
emacs /etc/init.d/irssid
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
#!/bin/bash

### BEGIN INIT INFO
# Provides:          irssid
# Required-Start:    $network
# Required-Stop:     $network
# Default-Start:     2 3 4 5
# Default-Stop:      0 1 6
# Short-Description: Start irssi daemon within screen session at boot time
# Description:       This init script will start an irssi session under screen using the settings provided in /etc/irssid.conf
### END INIT INFO

# Include the LSB library functions
. /lib/lsb/init-functions

# Setup static variables
configFile='/etc/irssid.conf'
daemonExec='/usr/bin/screen'
daemonArgs='-D -m'
daemonName="$(basename "$daemonExec")"
pidFile='/var/run/irssid.pid'

#
# Checks if the environment is capable of running the script (such as
# availability of programs etc).
#
# Return: 0 if the environmnt is properly setup for execution of init script, 1
#         if not all conditions have been met.
#
function checkEnvironment() {
    # Verify that the necessary binaries are available for execution.
    local binaries=(irssi screen)

    for bin in "${binaries[@]}"; do
        if ! which "$bin" > /dev/null; then
            log_failure_msg "Binary '$bin' is not available. Please install \
package containing it."
            exit 5
        fi
    done
}

#
# Checks if the configuration files are available and properly setup.
#
# Return: 0 if irssid if properly configured, 1 otherwise.
#
function checkConfig() {
    # Make sure the configuration file has been created
    if ! [[ -f $configFile ]]; then
        log_failure_msg "Please populate the configuration file '$configFile' \
before running."
        exit 6
    fi

    # Make sure the required options have been set
    local reqOptions=(user group session)
    for option in "${reqOptions[@]}"; do
        if ! grep -q -e "^[[:blank:]]*$option=" "$configFile"; then
            log_failure_msg "Mandatory option '$option' was not specified in \
'$configFile'"
            exit 6
        fi
    done
}

#
# Loads the configuration file and performs any additional configuration steps.
#
function configure() {
    . "$configFile"
    daemonArgs="$daemonArgs -S $session irssi"
    [[ -n $args ]] && daemonArgs="$daemonArgs $args"
    daemonCommand="$daemonExec $daemonArgs"
}

#
# Starts the daemon.
#
# Return: LSB-compliant code.
#
function start() {
    start-stop-daemon --start -v -b -x /bin/su -p /tmp/irssi.screen.session -m --chdir /home/$user -- - $user -c "screen -D -m -S irssi -- irssi" 1>>/log.irssi
}

#
# Stops the daemon.
#
# Return: LSB-compliant code.
#
function stop() {
    start-stop-daemon --stop -x /bin/su -p /tmp/irssi.screen.session -q
}

checkEnvironment
checkConfig
configure

case "$1" in
    start)
        log_daemon_msg "Starting daemon" "irssid"
        start && log_end_msg 0 || log_end_msg $?
        ;;
    stop)
        log_daemon_msg "Stopping daemon" "irssid"
        stop && log_end_msg 0 || log_end_msg $?
        ;;
    restart)
        log_daemon_msg "Restarting daemon" "irssid"
        stop
        start && log_end_msg 0 || log_end_msg $?
        ;;
    force-reload)
        log_daemon_msg "Restarting daemon" "irssid"
        stop
        start && log_end_msg 0 || log_end_msg $?
        ;;
    status)
        status_of_proc -p "$pidFile" "$daemonExec" screen && exit 0 || exit $?
        ;;
    *)
        echo "irssid (start|stop|restart|force-reload|status|help)"
        ;;
esac
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod +x /etc/init.d/irssid
#+END_SRC

Create a configuration file, replacing /myusername/ with your username.

#+BEGIN_SRC: bash
emacs /etc/irssid.conf
#+END_SRC

#+BEGIN_SRC: bash
#
# Configuration file for irssid init script
#
# Mandatory options:
#
#    user    - Specify user for running irssi.
#    group   - Specify group for running irssi.
#    session - Specify screen session name to be used for irssi.
#
# Non-mandatory options:
#
#    args    - Pass additional arguments to irssi.
#

user='myusername'
group='irssi'
session='irssi'
args='--config /home/myusername/.irssi/config'
#+END_SRC

Save and exit.  Then add your user to the irssi group and start the daemon.

#+BEGIN_SRC: bash
groupadd irssi
usermod -aG irssi myusername
update-rc.d irssid defaults
chown -R myusername:irssi /home/myusername/.irssi
service irssid start
#+END_SRC

Create a script to make running IRC on the server easier.

#+BEGIN_SRC: bash
emacs /usr/bin/irc
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
#!/bin/bash
screen -r irssi
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod +x /usr/bin/irc
chown myusername:myusername /usr/bin/irc
#+END_SRC

Then to subsequently access irssi log into the BBB using ssh and type:

#+BEGIN_SRC: bash
irc
#+END_SRC

** Install a Jabber/XMPP server

#+BEGIN_VERSE
/Well heck, it isn’t that hard to write an instant messaging system./

--Jeremie Miller
#+END_VERSE

*** The Server

Generate a SSL certificate.

#+BEGIN_SRC: bash
openssl ecparam -out /etc/ssl/private/xmpp.pem -name prime256v1
openssl genpkey -paramfile /etc/ssl/private/xmpp.pem -out /etc/ssl/private/xmpp.key
openssl req -new -x509 -key /etc/ssl/private/xmpp.key -out /etc/ssl/certs/xmpp.crt -days 3650
#+END_SRC

The above uses a Diffie-Hellman elliptic curve (ECDH P-256) algorithm. It is apparent that amongst crypographers there are differences of opinion about the security of elliptic curves, so if you prefer there is also a more traditional RSA way to generate an SSL certificate:

#+BEGIN_SRC: bash
openssl genrsa -out /etc/ssl/private/xmpp.key 4096
openssl req -new -x509 -key /etc/ssl/private/xmpp.key -out /etc/ssl/certs/xmpp.crt -days 3650
#+END_SRC

Change permissions.

#+BEGIN_SRC: bash
chmod 600 /etc/ssl/private/xmpp.key
chmod 600 /etc/ssl/certs/xmpp.crt
chown prosody:prosody /etc/ssl/private/xmpp.key
chown prosody:prosody /etc/ssl/certs/xmpp.crt
#+END_SRC

Install Prosody.

#+BEGIN_SRC: bash
apt-get install prosody
cp -a /etc/prosody/conf.avail/example.com.cfg.lua /etc/prosody/conf.avail/xmpp.cfg.lua
emacs /etc/prosody/conf.avail/xmpp.cfg.lua
#+END_SRC

Change the *VirtualHost* name to your domain name and remove the line below it.

Set the ssl section to:

#+BEGIN_SRC: bash
	ssl = {
		key = "/etc/ssl/private/xmpp.key";
		certificate = "/etc/ssl/certs/xmpp.crt";
		}
#+END_SRC

And also append the following:

#+BEGIN_SRC: bash
modules_enabled = {
  "bosh"; -- Enable mod_bosh
  "tls"; -- Enable mod_tls
}

c2s_require_encryption = true
s2s_require_encryption = true
#+END_SRC

Save and exit.  Create a symbolic link.

#+BEGIN_SRC: bash
ln -sf /etc/prosody/conf.avail/xmpp.cfg.lua /etc/prosody/conf.d/xmpp.cfg.lua
#+END_SRC

Add a user.  You will be prompted to specify a password.  You can repeat the process for as many users as needed.  This will also be your Jabber ID (JID).

#+BEGIN_SRC: bash
prosodyctl adduser myusername@mydomainname.com
#+END_SRC

Restart the server

#+BEGIN_SRC: bash
service prosody restart
#+END_SRC

On your internet router/firewall open ports 5222, 5223, 5269, 5280 and 5281 and forward them to the BBB.

It's possible to test that your XMPP server is working at https://xmpp.net.  It may take several minutes and you'll get a low score because of the self-signed certificate, but it will at least verify that your server is capable of communicating.

*** Managing users

To add a user:

#+BEGIN_SRC: bash
prosodyctl adduser myusername@mydomainname.com
#+END_SRC

To change a user password:

#+BEGIN_SRC: bash
prosodyctl passwd myusername@mydomainname.com
#+END_SRC

To remove a user:

#+BEGIN_SRC: bash
prosodyctl deluser myusername@mydomainname.com
#+END_SRC

Report the status of the XMPP server:

#+BEGIN_SRC: bash
prosodyctl status
#+END_SRC

*** Using with Jitsi
Jitsi is the recommended communications client for desktop or laptop systems, since it includes the /off the record/ (OTR) feature which provides some additional security beyond the usual SSL certificates.

Jitsi can be downloaded from https://jitsi.org/

On your desktop/laptop open Jitsi and select *Options* from the *Tools* menu.

Click *Add* to add a new user, then enter the Jabber ID which you previously specified with /prosodyctl/ when setting up the XMPP server. Close and then you should notice that your status is "Online" (or if not then you should be able to set it to online).

From the *File* menu you can add contacts, then select the chat icon to begin a chat.  Click on the lock icon on the right hand side and this will initiate an authentication procedure in which you can specify a question and answer to verify the identity of the person you're communicating with.  Once authentication is complete then you'll be chating using OTR, which provides an additional layer of security.

When opening Jitsi initially you will get a certificate warning for your domain name (assuming that you're using a self-signed certificate). If this happens then select *View Certificate* and enable the checkbox to trust the certificate, then select *Continue Anyway*.  Once you've done this then the certificate warning will not appear again unless you reinstall Jitsi or use a different computer.

You can also [[https://www.youtube.com/watch?v=vgx7VSrDGjk][see this video]] as an example of using OTR.
*** Using with Ubuntu
The default XMPP client in Ubuntu is Empathy.  Using Empathy isn't as secure as using Jitsi, since it doesn't include the /off the record/ feature, but since it's the default it's what many users will have easy access to.

Open *System Settings* and select *Online Accounts*, *Add account*  and then *Jabber*.

Enter your username (myusername@mydomainname.com) and password.

Click on *Advanced* and make sure that *Encryption required* and *Ignore SSL certificate errors* are checked.  Ignoring the certificate errors will allow you to use the self-signed certificate created earlier.  Then click *Done* and set your Jabber account and Empathy to *On*.
*** Using with Android

There are a few XMPP clients available on Android.  Ideally choose ones which support off-the-record messaging.  Here are some examples.

**** Xabber
Install [[https://f-droid.org/][F-Droid]]

Search for and install Xabber.

Add an account and enter your Jabber/XMPP ID and password.

From the menu select *Settings* then *Security* then *OTR mode*.  Set the mode to *Required*.

Make sure that *Check server certificate* is not checked.

Go back to the initial screen and then using the menu you can add contacts and begin chatting.  Both parties will need to go through the off-the-record question and answer verification before the chat can begin, but that only needs to be done once for each person you're chatting with.

**** Gibberbot
Install [[https://f-droid.org/][F-Droid]]

Search for and install Gibberbot, otherwise known as ChatSecure.

From the menu open *Accounts*

Select *Add account*

Change the server port from 0 to 5222

Done

Accept unknown certificate? Select *Always*

Go back to the initial screen and then using the menu you can add contacts and begin chatting.

** Social Networking

#+BEGIN_VERSE
/Facebook is not your friend, it is a surveillance engine./

-- Richard Stallman, Free Software Foundation
#+END_VERSE

*** Friendica
**** Installation

See [[Setting up a web site]] for details of how to update the Apache configuration for your Friendica site.  You should have a separate domain name specifically to run Friendica on.  It can't be installed in a subdirectory on a domain used for something else.

Edit your Apache configuration and disable the port 80 (HTTP) version of the site.  We only want to log into Friendica via HTTPS, so to prevent anyone from accidentally logging in insecurely:

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/mydomainname.com
#+END_SRC

Within the section which begins with *<VirtualHost *:80>* change the following:

#+BEGIN_SRC: bash
  <Directory /var/www/mydomainname.com/htdocs/>
    deny from all
  </Directory>
#+END_SRC

Save and exit, then restart the apache server.

#+BEGIN_SRC: bash
service apache2 restart
#+END_SRC

Now install some dependencies.

#+BEGIN_SRC: bash
apt-get install mysql-server php5-common php5-cli php5-curl php5-gd php5-mysql php5-mcrypt
#+END_SRC

Enter an admin password for MySQL.

Reduce the memory use of mysql by using the "small" configuration.

#+BEGIN_SRC: bash
cp /usr/share/doc/mysql-server-5.5/examples/my-small.cnf /etc/mysql/my.cnf
#+END_SRC

Create a mysql database.

#+BEGIN_SRC: bash
mysql -u root -p
create database friendica;
CREATE USER 'friendicaadmin'@'localhost' IDENTIFIED BY 'password';
GRANT ALL PRIVILEGES ON friendica.* TO 'friendicaadmin'@'localhost';
quit
#+END_SRC

You may need to fix Git SSL problems.

#+BEGIN_SRC: bash
git config --global http.sslVerify true
apt-get install ca-certificates
cd ~/
emacs .gitconfig
#+END_SRC

The .gitconfig file should look something like this:

#+BEGIN_SRC: bash
[http]
        sslVerify = true
        sslCAinfo = /etc/ssl/certs/ca-certificates.crt
[user]
        email = myusername@mydomainname.com
        name = yourname
#+END_SRC

Get the source code.

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
cd /var/www/$HOSTNAME
mv htdocs htdocs_old
git clone https://github.com/friendica/friendica.git htdocs
chmod -R 755 htdocs
chown -R www-data:www-data htdocs
chown -R www-data:www-data htdocs/view/smarty3
git clone https://github.com/friendica/friendica-addons.git htdocs/addon
#+END_SRC

Now visit the URL of your site and you should be taken through the rest of the installation procedure.  If you have trouble with "allow override" ensure that "AllowOverride" is set to "all" in your Apache settings for the site (within /etc/apache2/sites-available) and then restart the apache2 service.

Install the poller.

#+BEGIN_SRC: bash
emacs /etc/crontab
#+END_SRC

and append the following, changing mydomainname.com to whatever your domain is.

#+BEGIN_SRC: bash
*/10 *  * * *	root	cd /var/www/mydomainname.com/htdocs; /usr/bin/timeout 120 /usr/bin/php include/poller.php
#+END_SRC

Save and exit, then restart cron.

#+BEGIN_SRC: bash
service cron restart
#+END_SRC

You can improve the speed of Friendica database searches by adding the following indexes:

#+BEGIN_SRC: bash
mysql -u root -p
use friendica;
CREATE INDEX `uri_received` ON item(`uri`, `received`);
CREATE INDEX `received_uri` ON item(`received`, `uri`);
CREATE INDEX `contact-id_created` ON item(`contact-id`, created);
CREATE INDEX `uid_network_received` ON item(`uid`, `network`, `received`);
CREATE INDEX `uid_parent` ON item(`uid`, `parent`);
CREATE INDEX `uid_received` ON item(`uid`, `received`);
CREATE INDEX `uid_network_commented` ON item(`uid`, `network`, `commented`);
CREATE INDEX `uid_title` ON item(uid, `title`);
CREATE INDEX `created_contact-id` ON item(`created`, `contact-id`);
quit
#+END_SRC

Make sure that Friendica doesn't use too much memory.

#+BEGIN_SRC: bash
emacs /var/www/$HOSTNAME/htdocs/.htaccess
#+END_SRC

Append the following:

#+BEGIN_SRC: bash
php_value memory_limit 32M
#+END_SRC

The save ane exit.

**** Backups

Make sure that the database gets backed up.  By using cron if anything goes wrong then you should be able to recover the database either from the previous day or the previous week.

#+BEGIN_SRC: bash
emacs /etc/cron.daily/friendicabackup
#+END_SRC

Enter the following

#+BEGIN_SRC: bash
#!/bin/sh

MYSQL_PASSWORD=<mysql root password>

umask 0077

# Backup the database
mysqldump --password=$MYSQL_PASSWORD friendica > /var/backups/friendica_daily.sql

# Make the backup readable only by root
chmod 600 /var/backups/friendica_daily.sql
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod 600 /etc/cron.daily/friendicabackup
chmod +x /etc/cron.daily/friendicabackup
emacs /etc/cron.weekly/friendicabackup
#+END_SRC

Enter the following

#+BEGIN_SRC: bash
#!/bin/sh

MYSQL_PASSWORD=<mysql root password>

umask 0077

# Backup the database
mysqldump --password=$MYSQL_PASSWORD friendica > /var/backups/friendica_weekly.sql

# Make the backup readable only by root
chmod 600 /var/backups/friendica_weekly.sql
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod 600 /etc/cron.weekly/friendicabackup
chmod +x /etc/cron.weekly/friendicabackup
#+END_SRC

**** Recommended configuration
***** Admin
To get to the admin settings you will need to be logged in with the admin email address which you specified at the beginning of the installation procedure.  Depending upon the theme which you're using "/admin/" will be available either as an icon or on a drop down menu.

Under the *plugins* section the main one which you may wish to enable is the NSFW plugin.  With that enabled if a post contans the #NSFW tag then it will appear minimised by default and you will need to click a button to open it.

Under the *themes* section select a few themes, including mobile themes which are suitable for phones or tablets.

Under the *site* section give your Friendica node a name other than "/my friend network/", you can change the icon and banner text and set the default mobile theme typically to /frost-mobile/.  If you don't want your node to host a lot of accounts for people you don't know then you may want to set the register policy to "/requires approval/".  For security it's probably a good idea only to host accounts for people who you actually know, rather than random strangers.  Also be aware that the Beaglebone does not have a great deal of computational power or bandwidth and will not function well if there are hundreds of users using your node.  If you're not federating with Diaspora or other sites then you may wish to select "/only allow Friendica contacts/".  That improves the security of the system, since communication between Friendica nodes is always encrypted separately and in addition to the usual SSL encryption layer - which makes life interesting for the Surveillance State and at least keeps those cryptanalysts employed.

It's probably a good idea to enable "/private posts by default for new users/" and also "/don't include post content in email notifications/". Since traditional email isn't a secure system and is easily vulnerable to attack by systems such as [[https://en.wikipedia.org/wiki/XKeyscore][Xkeyscore]].

***** Settings
Each user has their own customisable settings, typically available either via an icon or by an entry on a drop down menu.

Under *additional features* enable "/richtext editor/", "/post preview/", "/group filter/", "/network filter/", "/edit sent posts/" and "/dislike posts/".

Under *display settings* select your desktop and mobile themes.

Once you have connected to enough friends it's also a good idea to use the "/export personal data/" option from here.  This will save a file to your local system, which you can import into another friendica node if necessary.
**** To access from an Android device
***** App
Open a browser on your device and go to https://f-droid.org/ then download and install the F-Droid apk.  If you then open F-Droid you can search for and install the Friendica app.

If you are using a self-signed certificate then at the login screen scroll down to the bottom, select the SSL settings then scroll down and disable SSL certificate checks.  You will then be able to log in using https, which at least gives you some protection via the encryption.

More information about the Friendica app can be found on http://friendica-for-android.wiki-lab.net/
***** Mobile Theme
Another way to access Friendica from a mobile device is to just use the web browser.  If you have selected a mobile theme within your settings then when viewing from an Android system the mobile theme will be displayed.
*** Movim

#+BEGIN_VERSE
/The way we communicate with others and with ourselves ultimately determines the quality of our lives/

-- Anthony Robbins
#+END_VERSE

Movim is another social networking system based around the XMPP protocol.

You will need to have previously [[Install a Jabber/XMPP server][installed the Jabber/XMPP server]].

Edit your Apache configuration and disable the port 80 (HTTP) version of the site.  We only want to log into Movim via HTTPS, so to prevent anyone from accidentally logging in insecurely:

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/mydomainname.com
#+END_SRC

Within the section which begins with *<VirtualHost *:80>* add the following:

#+BEGIN_SRC: bash
  <Directory /var/www/mydomainname.com/htdocs/movim>
    deny from all
  </Directory>
#+END_SRC

Within the section which begins with *<VirtualHost *:443>* add the following:

#+BEGIN_SRC: bash
  <Directory /var/www/mydomainname.com/htdocs/movim>
    Options Indexes FollowSymLinks MultiViews
    AllowOverride All
    Order allow,deny
    allow from all
  </Directory>
#+END_SRC

Save and exit, then restart the apache server.

#+BEGIN_SRC: bash
service apache2 restart
#+END_SRC

Download the source.

#+BEGIN_SRC: bash
cd /tmp
wget http://freedombone.uk.to/movim.tar.gz
#+END_SRC

Verify it.

#+BEGIN_SRC: bash
sha256sum movim.tar.gz
2740ddbedf6cefcc2934759374376643b6cdea4fb7f944ec25098a6868cb499e movim.tar.gz
#+END_SRC

Install it.

#+BEGIN_SRC: bash
tar -xzvf movim.tar.gz
export HOSTNAME=mydomainname.com
cp -r movim-* /var/www/$HOSTNAME/htdocs/movim
chmod 755 /var/www/$HOSTNAME/htdocs/movim
chown -R www-data:www-data /var/www/$HOSTNAME/htdocs/movim
#+END_SRC

Install some MySql prerequisites.

#+BEGIN_SRC: bash
apt-get install mysql-server php5-common php5-cli php5-curl php5-gd php5-mysql php5-mcrypt
#+END_SRC

If necessary, enter an admin password for MySQL.

Reduce the memory use of mysql by using the "small" configuration.

#+BEGIN_SRC: bash
cp /usr/share/doc/mysql-server-5.5/examples/my-small.cnf /etc/mysql/my.cnf
#+END_SRC

Create a mysql database.

#+BEGIN_SRC: bash
mysql -u root -p
create database movim;
CREATE USER 'movimadmin'@'localhost' IDENTIFIED BY 'movimadminpassword';
GRANT ALL PRIVILEGES ON movim.* TO 'movimadmin'@'localhost';
quit
#+END_SRC

With a web browser navigate to:

https://mydomainname.com/movim/admin

Enter /admin/ as the username and /password/ as the password.

Click on /General Settings/ and alter the administrator username to /movimadmin/ and password to some long random string (using a password manager such as KeepassX).

Change the /Environment/ from /Development/ to /Production/.

The /BOSH URL/ should be http://localhost:5280/http-bind (TODO: should this be https://localhost:5281/http-bind and if so do certificate warnings need to be disabled?)

Click /Submit/ followed by /Resend/.

Click on /Database Settings/ and alter the MySql movim database username to /movimadmin/ and password to the password you specified in the previous step.

Click /Submit/ followed by /Resend/.  If you get a lot of orange warnings about database fields being created then hit /Submit/ again until you see "Movim database is up to date".

If everything on all three tabs looks green then you are ready to go. Click on the Movim logo at the top left and then log in with your Jabber ID (JID).

*** Red Matrix
**** Introduction
Red Matrix is the current version of the Friendica social networking system.  It's more general than Friendica in that it's designed as a generic communication system based around a protocol called "zot".  At the time of writing in early 2014 Red Matrix remains at an alpha stage of development and so it's not advised that you install it unless you're willing to put up with bugs and frustrations.  In the large majority of cases it's better to stick with Friendica for now.

**** Prerequisites
The main problem with Red Matrix is that in order to install it you will need to have purchased a domain name (i.e. not a FreeDNS subdomain) and a SSL certificate for it.

You could join some other Red Matrix server, but this suffers from "/The Levison Problem/" in which some goons show up with a gagging order demanding coppies of the SSL private key.  In that scenario unless the owner of the server is exceptionally brave users may never be informed that the site has been compromised or that there is interception hardware attached to the server.  Joining another server defeats the object of being digitally self-sufficient and raises legal question marks about the ownership of data which you might upload to a server which doesn't belong to you.

**** Installation

See [[Setting up a web site]] for details of how to update the Apache configuration for your Red Matrix site.  You should have a separate domain name specifically to run Red Matrix on.  It can't be installed in a subdirectory on a domain used for something else.

Edit your Apache configuration and disable the port 80 (HTTP) version of the site.  We only want to log into Red Matrix via HTTPS, so to prevent anyone from accidentally logging in insecurely:

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/mydomainname.com
#+END_SRC

Replace the section which begins with *<VirtualHost *:80>* with the following:

#+BEGIN_SRC: bash
<VirtualHost *:80>
	ServerAdmin myusername@mydomainname.com
	ServerName myredmatrixdomainname.com

    RewriteEngine On
    RewriteCond %{HTTPS} off
    RewriteRule (.*) https://%{HTTP_HOST}%{REQUEST_URI}
</VirtualHost>
#+END_SRC

Save and exit, then restart the apache server.

#+BEGIN_SRC: bash
service apache2 restart
#+END_SRC

Now install some dependencies.

#+BEGIN_SRC: bash
apt-get install mysql-server php5-common php5-cli php5-curl php5-gd php5-mysql php5-mcrypt
#+END_SRC

Enter an admin password for MySQL.

Reduce the memory use of mysql by using the "small" configuration.

#+BEGIN_SRC: bash
cp /usr/share/doc/mysql-server-5.5/examples/my-small.cnf /etc/mysql/my.cnf
#+END_SRC

Create a mysql database.

#+BEGIN_SRC: bash
mysql -u root -p
create database redmatrix;
CREATE USER 'redmatrixadmin'@'localhost' IDENTIFIED BY 'password';
GRANT ALL PRIVILEGES ON redmatrix.* TO 'redmatrixadmin'@'localhost';
quit
#+END_SRC

You may need to fix Git SSL problems.

#+BEGIN_SRC: bash
git config --global http.sslVerify true
apt-get install ca-certificates
cd ~/
emacs .gitconfig
#+END_SRC

The .gitconfig file should look something like this:

#+BEGIN_SRC: bash
[http]
        sslVerify = true
        sslCAinfo = /etc/ssl/certs/ca-certificates.crt
[user]
        email = myusername@mydomainname.com
        name = yourname
#+END_SRC

Get the source code.

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
cd /var/www/$HOSTNAME
rm -rf htdocs
git clone https://github.com/friendica/red.git htdocs
chmod -R 755 htdocs
chown -R www-data:www-data htdocs
mkdir htdocs/view/tpl/smarty3
chmod 777 htdocs/view/tpl
chmod 777 htdocs/view/tpl/smarty3
git clone https://github.com/friendica/red-addons.git htdocs/addon
#+END_SRC

Now visit the URL of your site and you should be taken through the rest of the installation procedure.  Note that this may take a few minutes so don't be concerned if it looks as if it has crashed - just leave it running.  If you have trouble with "allow override" ensure that "AllowOverride" is set to "all" in your Apache settings for the site (within /etc/apache2/sites-available) and then restart the apache2 service.

Install the poller.

#+BEGIN_SRC
emacs /etc/crontab
#+END_SRC

and append the following, changing mydomainname.com to whatever your domain is.

#+BEGIN_SRC
12,22,32,42,52 * *   *   *   root cd /var/www/apespace.org/htdocs; /usr/bin/timeout 240 /usr/bin/php include/poller.php
#+END_SRC

Save and exit, then restart cron.

#+BEGIN_SRC: bash
service cron restart
#+END_SRC

**** Backups

Make sure that the database gets backed up.  By using cron if anything goes wrong then you should be able to recover the database either from the previous day or the previous week.

#+BEGIN_SRC: bash
emacs /etc/cron.daily/redmatrixbackup
#+END_SRC

Enter the following

#+BEGIN_SRC: bash
#!/bin/sh

MYSQL_PASSWORD=<mysql root password>

umask 0077

# Backup the database
mysqldump --password=$MYSQL_PASSWORD redmatrix > /var/backups/redmatrix_daily.sql

# Make the backup readable only by root
chmod 600 /var/backups/redmatrix_daily.sql
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod 600 /etc/cron.daily/redmatrixbackup
chmod +x /etc/cron.daily/redmatrixbackup
emacs /etc/cron.weekly/redmatrixbackup
#+END_SRC

Enter the following

#+BEGIN_SRC: bash
#!/bin/sh

MYSQL_PASSWORD=<mysql root password>

umask 0077

# Backup the database
mysqldump --password=$MYSQL_PASSWORD redmatrix > /var/backups/redmatrix_weekly.sql

# Make the backup readable only by root
chmod 600 /var/backups/redmatrix_weekly.sql
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod 600 /etc/cron.weekly/redmatrixbackup
chmod +x /etc/cron.weekly/redmatrixbackup
#+END_SRC
**** To access from an Android device
***** App
Open a browser on your device and go to https://f-droid.org/ then download and install the F-Droid apk.  If you then open F-Droid you can search for and install the Friendica app.

If you are using a self-signed certificate then at the login screen scroll down to the bottom, select the SSL settings then scroll down and disable SSL certificate checks.  You will then be able to log in using https, which at least gives you some protection via the encryption.

More information about the Friendica app can be found on http://friendica-for-android.wiki-lab.net/

*** pump.io
pump.io is the successor to StatusNet (which later became [[GNU Social]]) and is a communications system which can do things other than just microblogging.  It takes fewer system resources to run and so is better suited to low power servers such as the BBB, but is more complicated to install.  Currently when using self-signed certificates it seems very hard to federate with other pump.io servers so it may be that although GNU Social is an older system it may still be more practical.  For the instructions which follow it will be possible to run your own pump.io site for your family and friends, as a kind of /data silo/, but federating with anyone else could turn out to be difficult or impossible.

A list of pump.io sites can be found at http://pumpstatus.jpope.org

For a pump.io site you will need a separate domain/subdomain, so see [[Setting up a web site]] for details of how to create an Apache configuration for your site.  If you're using freedns then you will need to create a new subdomain.

#+BEGIN_SRC: bash
apt-get update && apt-get install redis-server nodejs-legacy imagemagick graphicsmagick git-core screen
cd /opt
git clone https://github.com/e14n/pump.io.git
cd /opt/pump.io
npm install
npm install databank-redis
#+END_SRC

Edit the configuration file.

#+BEGIN_SRC: bash
emacs /etc/pump.io.json
#+END_SRC

Add the following, replacing /mypumpiodomainname.com/ with your domain name.

#+BEGIN_SRC: bash
{
    "driver": "redis",
    "params": {"host":"localhost","port":6379},
    "secret":  "A long random string",
    "noweb":  false,
    "site":  "Name of my pump.io site",
    "owner":  "My name or organisation",
    "ownerURL":  "https://mypumpiodomainname.com/",
    "port":  7270,
    "urlPort": 443,
    "hostname":  "mypumpiodomainname.com",
    "address":  "localhost",
    "nologger":  false,
    "serverUser":  "pumpio",
	"rejectUnauthorized": false,
    "key":  "/var/local/pump.io/keys/mypumpiodomainname.com.key",
    "cert":  "/var/local/pump.io/keys/mypumpiodomainname.com.crt",
    "uploaddir": "/var/local/pump.io/uploads",
    "debugClient": false,
    "firehose": "ofirehose.example",
	"logfile": "/var/local/pump.io/pump.io.log",
    "disableRegistration": false
}
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
export HOSTNAME=mypumpiodomainname.com
mkdir /var/local/pump.io
mkdir /var/local/pump.io/uploads
mkdir /var/local/pump.io/keys
cp /etc/ssl/private/$HOSTNAME.key /var/local/pump.io/keys
cp /etc/ssl/certs/$HOSTNAME.crt /var/local/pump.io/keys
useradd -s /bin/bash -d /var/local/pump.io pumpio
chown -R pumpio:pumpio /var/local/pump.io
chmod 400 /var/local/pump.io/keys/*
mkdir /tmp/apache2
cd /tmp/apache2
apt-get build-dep apache2
apt-get install autoconf
apt-get source apache2
cd apache2-*
wget http://freedombone.uk.to/apache-2.2-wstunnel.patch
sha256sum apache-2.2-wstunnel.patch
cfc4866da2688a8eb76e0300cf16b52539ef4e525053a3851d4b6bba9a77e439

patch -p1 -i apache-2.2-wstunnel.patch
autoconf
./configure --enable-so --enable-proxy=shared --enable-proxy-wstunnel=shared
make
cp modules/proxy/.libs/mod_proxy_wstunnel.so  /usr/lib/apache2/modules/
cd /etc/apache2/mods-enabled
ln -s /usr/lib/apache2/modules/mod_proxy_wstunnel.so ../mods-available/proxy_wstunnel.load
#+END_SRC

Within the section of your Apache site configuration:

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/mypumpiodomainname.com
#+END_SRC

The initial section which begins with *<VirtualHost *:80>* should be replaced by the following, replacing /mypumpiodomainname.com/ with your pump.io domain name and /myusername@mydomainname.com/ with your email address.

#+BEGIN_SRC: bash
<VirtualHost *:80>
	ServerAdmin myusername@mydomainname.com
	ServerName mypumpiodomainname.com

    RewriteEngine On
    RewriteCond %{HTTPS} off
    RewriteRule (.*) https://%{HTTP_HOST}%{REQUEST_URI}
</VirtualHost>
#+END_SRC

Add the following in the section which begins with *<VirtualHost *:443>*.

#+BEGIN_SRC: bash
    LoadModule proxy_wstunnel_module /usr/lib/apache2/modules/mod_proxy_wstunnel.so

    <Location /main/realtime/sockjs>
        ProxyPass wss://localhost/main/realtime/sockjs
        ProxyPassReverse wss://localhost/main/realtime/sockjs
    </Location>

#    <LocationMatch ".*\.(jpg|png|gif)$">
#        CacheEnable disk
#    </LocationMatch>

    ProxyVia On
    ProxyPreserveHost On
    SSLProxyEngine On

    ProxyPass / https://localhost:7270/
    ProxyPassReverse / https://localhost:7270/
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
a2enmod cache
a2enmod disk_cache
apachectl configtest
service apache2 restart
npm install forever -g
#+END_SRC

Now create the daemon.

#+BEGIN_SRC: bash
emacs /etc/init.d/pumpio
#+END_SRC

Add the following text:

#+BEGIN_SRC: bash
#!/bin/bash
# /etc/init.d/pumpio

### BEGIN INIT INFO
# Provides:          pump.io
# Required-Start:    $remote_fs $syslog
# Required-Stop:     $remote_fs $syslog
# Default-Start:     2 3 4 5
# Default-Stop:      0 1 6
# Short-Description: starts pump.io as a background daemon
# Description:       Starts pump.io on boot
### END INIT INFO

# Author: Bob Mottram <bob@robotics.uk.to>

#Settings
SERVICE='pumpio'
COMMAND="forever /opt/pump.io/bin/pump > /var/local/pump.io/daemon.log"
USERNAME='pumpio'
NICELEVEL=19 # from 0-19 the bigger the number, the less the impact on system resources
HISTORY=1024
INVOCATION="nice -n ${NICELEVEL} ${COMMAND}"
PATH='/usr/local/sbin:/usr/local/bin:/usr/bin:/sbin:/usr/sbin:/bin:/var/local/pump.io'

pumpio_start() {
echo "Starting $SERVICE..."
su --command "screen -h ${HISTORY} -dmS ${SERVICE} ${INVOCATION}" $USERNAME
}

pumpio_stop() {
echo "Stopping $SERVICE"
su --command "screen -p 0 -S ${SERVICE} -X stuff "'^C'"" $USERNAME
}

#Start-Stop here
case "$1" in
  start)
    pumpio_start
    ;;
  stop)
    pumpio_stop
    ;;
  restart)
    pumpio_stop
    sleep 10s
    pumpio_start
    ;;
    *)
  echo "Usage: $0 {start|stop|restart}"
  exit 1
  ;;
esac

exit 0
#+END_SRC

Save and exit. Then enable the daemon and run it.

#+BEGIN_SRC: bash
chmod +x /etc/init.d/pumpio
update-rc.d pumpio defaults
service pumpio start
#+END_SRC

Now visit your pump.io site by navigating to:

https://mypumpiodomainname.com

and add a new user.  If you wish this to be a single user node not open to the general public (including spammers and sockpuppets) then edit */etc/pump.io.json* and set *disableRegistration* to *true*.  After making that change restart with the command *service pumpio restart*.

** Install Gopher
*** Server setup

Gopher is an old internet protocol which originated a few years before the web and is purely text based.  It can be quite fun to build a gopher site and browse the gopherverse.  One thing to keep in mind is that there is no security with gopher, so any text transmitted is trivially interceptable by systems such as [[https://en.wikipedia.org/wiki/XKeyscore][Xkeyscore]] or deep packet inspection.

To set up a gopher server:

#+BEGIN_SRC: bash
apt-get install build-essential
cd /tmp
wget http://freedombone.uk.to/geomyidae-current.tgz
#+END_SRC

Verify the download:

#+BEGIN_SRC: bash
sha256sum geomyidae-current.tgz
162f55ab059ab0a9be8e840497795293bbd51c34b1f4564dcdf3f0ddd5c0db31 geomyidae-current.tgz
#+END_SRC

Then extract and install it.

#+BEGIN_SRC: bash
tar -xzvf geomyidae-current.tgz
cd geomyidae-*
make
make install
mkdir -p /var/gopher
#+END_SRC

Your content should be placed within /var/gopher with the index page being named index.gph.  The Gopher format is very simple - simpler than HTML - so creating pages is not much more difficult than editing a text file.

#+BEGIN_SRC: bash
emacs /etc/init.d/gopher
#+END_SRC

Enter the following:

#+BEGIN_SRC: bash
#! /bin/sh
### BEGIN INIT INFO
# Provides:          gopher
# Required-Start:    $remote_fs $syslog
# Required-Stop:     $remote_fs $syslog
# Default-Start:     2 3 4 5
# Default-Stop:      0 1 6
# Short-Description: Gopher daemon
# Description:       Gopher daemon
### END INIT INFO

# Do NOT "set -e"

# PATH should only include /usr/* if it runs after the mountnfs.sh script
PATH=/sbin:/usr/sbin:/bin:/usr/bin
DESC="Gopher daemon"
NAME=geomyidae
DAEMON=/usr/bin/$NAME
DAEMON_ARGS="-l /var/log/geomyidae.log -b /var/gopher -p 70"
PIDFILE=/var/run/$NAME.pid
SCRIPTNAME=/etc/init.d/$NAME

# Exit if the package is not installed
[ -x "$DAEMON" ] || exit 0

# Read configuration variable file if it is present
[ -r /etc/default/$NAME ] && . /etc/default/$NAME

# Load the VERBOSE setting and other rcS variables
. /lib/init/vars.sh

# Define LSB log_* functions.
# Depend on lsb-base (>= 3.2-14) to ensure that this file is present
# and status_of_proc is working.
. /lib/lsb/init-functions

#
# Function that starts the daemon/service
#
do_start()
{
	# Return
	#   0 if daemon has been started
	#   1 if daemon was already running
	#   2 if daemon could not be started
	start-stop-daemon --start --quiet --pidfile $PIDFILE --exec $DAEMON --test > /dev/null \
		|| return 1
	start-stop-daemon --start --quiet --pidfile $PIDFILE --exec $DAEMON -- \
		$DAEMON_ARGS \
		|| return 2
	# Add code here, if necessary, that waits for the process to be ready
	# to handle requests from services started subsequently which depend
	# on this one.  As a last resort, sleep for some time.
}

#
# Function that stops the daemon/service
#
do_stop()
{
	# Return
	#   0 if daemon has been stopped
	#   1 if daemon was already stopped
	#   2 if daemon could not be stopped
	#   other if a failure occurred
	start-stop-daemon --stop --quiet --retry=TERM/30/KILL/5 --pidfile $PIDFILE --name $NAME
	RETVAL="$?"
	[ "$RETVAL" = 2 ] && return 2
	# Wait for children to finish too if this is a daemon that forks
	# and if the daemon is only ever run from this initscript.
	# If the above conditions are not satisfied then add some other code
	# that waits for the process to drop all resources that could be
	# needed by services started subsequently.  A last resort is to
	# sleep for some time.
	start-stop-daemon --stop --quiet --oknodo --retry=0/30/KILL/5 --exec $DAEMON
	[ "$?" = 2 ] && return 2
	# Many daemons don't delete their pidfiles when they exit.
	rm -f $PIDFILE
	return "$RETVAL"
}

#
# Function that sends a SIGHUP to the daemon/service
#
do_reload() {
	#
	# If the daemon can reload its configuration without
	# restarting (for example, when it is sent a SIGHUP),
	# then implement that here.
	#
	start-stop-daemon --stop --signal 1 --quiet --pidfile $PIDFILE --name $NAME
	return 0
}

case "$1" in
  start)
	[ "$VERBOSE" != no ] && log_daemon_msg "Starting $DESC" "$NAME"
	do_start
	case "$?" in
		0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
		2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
	esac
	;;
  stop)
	[ "$VERBOSE" != no ] && log_daemon_msg "Stopping $DESC" "$NAME"
	do_stop
	case "$?" in
		0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
		2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
	esac
	;;
  status)
	status_of_proc "$DAEMON" "$NAME" && exit 0 || exit $?
	;;
  #reload|force-reload)
	#
	# If do_reload() is not implemented then leave this commented out
	# and leave 'force-reload' as an alias for 'restart'.
	#
	#log_daemon_msg "Reloading $DESC" "$NAME"
	#do_reload
	#log_end_msg $?
	#;;
  restart|force-reload)
	#
	# If the "reload" option is implemented then remove the
	# 'force-reload' alias
	#
	log_daemon_msg "Restarting $DESC" "$NAME"
	do_stop
	case "$?" in
	  0|1)
		do_start
		case "$?" in
			0) log_end_msg 0 ;;
			1) log_end_msg 1 ;; # Old process is still running
			*) log_end_msg 1 ;; # Failed to start
		esac
		;;
	  *)
		# Failed to stop
		log_end_msg 1
		;;
	esac
	;;
  *)
	#echo "Usage: $SCRIPTNAME {start|stop|restart|reload|force-reload}" >&2
	echo "Usage: $SCRIPTNAME {start|stop|status|restart|force-reload}" >&2
	exit 3
	;;
esac

:
#+END_SRC

Save and exit.  Then start the gopher service.

#+BEGIN_SRC: bash
chmod +x /etc/init.d/gopher
update-rc.d gopher defaults
service gopher start
#+END_SRC

On your internet router change the firewall settings to route port 70 to the BBB, then provided that you have a gopher plugin installed within your browser then you should be able to navigate to your gopher site with:

#+BEGIN_SRC: bash
gopher://mydomainname.com
#+END_SRC

There is a browser addon for Gopher called "overbite".  Installing that should enable you to view your site.

*** A phlogging script

A phlog is the gopher equivalent of a blog on the web.  You can create a script which makes phlogging easy.

#+BEGIN_SRC: bash
emacs /usr/bin/mkphlog
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
#!/bin/sh

# mkphlog - a utility to ease the creation of phlogs.
#           Organizes phlog posts in separate directories.
# Created by octotep; anyone can distribute, modify, and
# share this file however they please.
#
# Version 0.3
#
# Modified by Bob Mottram
#
# Please note, all date strings are in the form of mm/dd/yy(yy)

# The base of the entire gopher site.
gopherRoot="/var/gopher"

# The name of the phlog directory (contained in $gopherHome)
phlogDirName="phlog"

# Default editor, unless the user has one specified in env
editor=${EDITOR:-emacs}

# Default timezone, unless the user has one specified in env
TZ=${TZ:-UTC}

# Tells the script how many lines the title of the main page spans.
# Used to insert the newest post at the top.
# Titles created by mkphlog are 3 lines.
# Isn't used if $addTitleToMain is false
titleLineCount=3

entryDate=`date +%Y-%m-%d`

# Creates the phlog directory if it dosen't already exist.
CreatePhlogDir() {
    mkdir $phlogDirName
    chmod 755 $phlogDirName
    cd $phlogDirName
    echo "Phlog directory created."
}

# Updates the main phlog listing
UpdatePhlogListing() {
    # Just in case the user didn't specify a title
    if [ "$postTitleAns" = "" ] ; then
	echo -n "Do you want to create a blank post? (y/n) "
	read blankPostAns
	case $blankPostAns in
	    y* | Y* ) $postTitleAns="New Post" ;;
		    n* | N* ) echo "Goodbye, then." ; exit 1 ;;
			    * ) exit 1 ;;
				  esac
    fi

    cd $gopherRoot/$phlogDirName/
    title2=$(echo "${postTitleAns}" | tr " " _)
    postfilename="${entryDate}_${title2}.txt"
    touch ${postfilename}
    echo $postTitleAns >> ${postfilename}
    date "+%A %b %e %l:%M:%S %Y" >> ${postfilename}
    echo "------------------------------" >> ${postfilename}
    echo >> ${postfilename}
}


if [ -d $gopherRoot ] ; then
    cd $gopherRoot
else
    echo "You don't have a gopherspace set-up. Please run the gopher server setup instructions."
    exit 1
fi

if [ -d $phlogDirName ] ; then
    cd $phlogDirName
else
    echo -n "Do you want to create a phlog directory? (y/n) "
    read phlogDirAns
    case $phlogDirAns in
	y* | Y* ) CreatePhlogDir ;;
	n* | N* ) exit 1 ;;
	* ) exit 1 ;;
    esac
fi

echo -n "Would you like to create a phlog entry for today? (y/n) "
read phlogAns
case $phlogAns in
    y* | Y* ) echo "Creating today's phlog entry..." ;;
    n* | N* ) exit 0 ;;
    * ) exit 1 ;;
esac

# Make sure there isn't a post for that day, lest we overwrite it.
if [ ! -d $entryDate ]; then
    echo -n "Title: "
    read postTitleAns
    title2=$(echo "${postTitleAns}" | tr " " _)
    postfilename="${entryDate}_${title2}.txt"
    touch ${postfilename}
    chmod 644 ${postfilename}
    UpdatePhlogListing
    echo -n "Would you like to edit the post with $editor? (y/n) "
    read editorAns
    case $editorAns in
	y* | Y* ) $editor $gopherRoot/$phlogDirName/${postfilename} ;;
	n* | N* ) exit 0 ;;
	* ) exit 0 ;;
    esac

    rm $gopherRoot/$phlogDirName/${postfilename}~
else
    echo "There is already a post for today."
    echo -n "Would you like to edit the post with $editor? (y/n) "
    read editorAns
    case $editorAns in
	y* | Y* ) $editor $gopherRoot/$phlogDirName/$entryDate*.txt ;;
	n* | N* ) exit 0 ;;
	* ) exit 1 ;;
    esac
    rm $gopherRoot/$phlogDirName/${postfilename}.txt~
fi
exit 0
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod +x /usr/bin/mkphlog
#+END_SRC

Now entering the command /mkphlog/ will allow you to create a phlog entry.

** Install Owncloud

#+BEGIN_VERSE
/It's not water vapour/

-- Larry Ellison
#+END_VERSE

Owncloud will allow you to upload and download files, share photos, collaboratively edit documents, have a calendar and more.  You should be warned that Owncloud runs quite slowly via an ordinary web browser, but it can be a convenient way to access and share your data from any location in a reasonably secure manner.

*** Server Installation

Install some dependencies:

#+BEGIN_SRC: bash
apt-get install apache2 php5 php5-gd php-xml-parser php5-intl
apt-get install php5-sqlite php5-mysql smbclient curl libcurl3 php5-curl
#+END_SRC

It's very important that /mod_php5/ and not /mod_php5filter/ be installed.  If you have /mod_php5filter/ installed then Owncloud will always fail to install.

#+BEGIN_SRC: bash
a2dismod php5filter
apt-get install libapache2-mod-php5
#+END_SRC

Ensure that the size of files which may be uploaded or downloaded is large enough.

#+BEGIN_SRC: bash
emacs /etc/php5/apache2/php.ini
#+END_SRC

Set the following:

#+BEGIN_SRC: bash
upload_max_filesize = 512M
post_max_size = 512M
#+END_SRC

Save and exit, then edit your Apache configuration.

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
emacs /etc/apache2/sites-available/$HOSTNAME
#+END_SRC

And add the following, to the 443 VirtualHost section.  Really we only will want to be using Owncloud with HTTPS to ensure some level of security and avoidance of dragnet surveillance.

#+BEGIN_SRC: bash
  <Directory /var/www/mydomainname.com/htdocs/owncloud>
    Options Indexes FollowSymLinks MultiViews
    AllowOverride All
    Order allow,deny
    allow from all
  </Directory>
#+END_SRC

To ensure that nobody logs in insecurely add the following to the 80 VirtualHost section.

#+BEGIN_SRC: bash
  <Directory /var/www/mydomainname.com/htdocs/owncloud>
    deny from all
  </Directory>
#+END_SRC

Save and exit, then restart apache.

#+BEGIN_SRC: bash
service apache2 restart
#+END_SRC

Download owncloud.

#+BEGIN_SRC: bash
cd /tmp
wget http://freedombone.uk.to/owncloud.tar.bz2
#+END_SRC

Verify the download:

#+BEGIN_SRC: bash
sha256sum owncloud.tar.bz2
92b53fdfa7c4165b83dd2f8447f63928454a5815d08ff2d6165dd1a8969ecbe1 owncloud.tar.bz2
#+END_SRC

Extract the archive.  This may take a couple of minutes, so don't be alarmed that the system has crashed.

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
tar -xjf owncloud.tar.bz2
#+END_SRC

Move the extracted files to your site and set file permissions.

#+BEGIN_SRC: bash
cp -r owncloud /var/www/$HOSTNAME/htdocs
chown -R www-data:www-data /var/www/$HOSTNAME/htdocs/owncloud/apps
chown -R www-data:www-data /var/www/$HOSTNAME/htdocs/owncloud/config
chown www-data:www-data /var/www/$HOSTNAME/htdocs/owncloud
#+END_SRC

Edit the htaccess file for Owncloud.

#+BEGIN_SRC: bash
emacs /var/www/$HOSTNAME/htdocs/owncloud/.htaccess
#+END_SRC

Set the following.

#+BEGIN_SRC: bash
php_value upload_max_filesize 512M
php_value post_max_size 512M
php_value memory_limit 32M
#+END_SRC

Save and exit.

With a web browser visit your domain (mydomainname.com/owncloud) and enter an administrator username and password.

*** Owncloud on Android

First install [[https://f-droid.org/][F-Droid]] and then search for the current Owncloud app.  Once it's installed you'll then be able to log into the BBB with the URL https://mydomainname.com/opencloud, supplying your username and password.

** Install a Wiki

#+BEGIN_VERSE
/I believe that technology can liberate, but you need to be a master rather than a user. You need to pull technology apart and master it rather than letting it control you./

-- Tom Barbalet
#+END_VERSE

Dokuwiki is based upon flat files, and so is easy to move from one server to another without a lot of database complications.

Download the wiki.

#+BEGIN_SRC: bash
cd /tmp
wget http://freedombone.uk.to/dokuwiki.tgz
#+END_SRC

Verify it.

#+BEGIN_SRC: bash
sha256sum dokuwiki.tgz
6b126f90979463d9ddaa74acc6f96aa230cfdc789946f241c3646086d9574be8 dokuwiki.tgz
#+END_SRC

Then extract and install it.

#+BEGIN_SRC: bash
export HOSTNAME=mywikidomainname.com
tar -xzvf dokuwiki.tgz
mv /var/www/$HOSTNAME/htdocs /var/www/$HOSTNAME/htdocs_old
mv dokuwiki /var/www/$HOSTNAME/htdocs
#+END_SRC

Edit the Apache configuration for your wiki site.

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/$HOSTNAME
#+END_SRC

The settings should look something like the following.  Replace /mywikidomainname.com/ with your wiki domain name.

#+BEGIN_SRC: bash
<VirtualHost *:80>
	ServerAdmin myusername@mywikidomainname.com
	ServerName mydomainname.com

	DocumentRoot /var/www/mywikidomainname.com/htdocs

    <Directory /var/www/mywikidomainname.com/htdocs>
        order deny,allow
        allow from all
    </Directory>

    <LocationMatch "/(data|conf|bin|inc)/">
        order allow,deny
        deny from all
        satisfy all
    </LocationMatch>

	<Directory />
		Options FollowSymLinks
		AllowOverride All
	</Directory>

	ScriptAlias /cgi-bin/ /usr/lib/cgi-bin/
	<Directory "/usr/lib/cgi-bin">
		AllowOverride All
		Options +ExecCGI -MultiViews +SymLinksIfOwnerMatch
		Order allow,deny
		Allow from all
	</Directory>

	ErrorLog ${APACHE_LOG_DIR}/error.log

	# Possible values include: debug, info, notice, warn, error, crit,
	# alert, emerg.
	LogLevel error

	CustomLog ${APACHE_LOG_DIR}/access.log combined
</VirtualHost>

<IfModule mod_ssl.c>
<VirtualHost *:443>
	ServerAdmin myusername@mywikidomainname.com
	ServerName mywikidomainname.com

	DocumentRoot /var/www/mywikidomainname.com/htdocs

    <Directory /var/www/mywikidomainname.com/htdocs>
        order deny,allow
        allow from all
    </Directory>

    <LocationMatch "/(data|conf|bin|inc)/">
        order allow,deny
        deny from all
        satisfy all
    </LocationMatch>

	<Directory />
		Options FollowSymLinks
		AllowOverride All
	</Directory>

	ScriptAlias /cgi-bin/ /usr/lib/cgi-bin/
	<Directory "/usr/lib/cgi-bin">
		AllowOverride All
		Options +ExecCGI -MultiViews +SymLinksIfOwnerMatch
		Order allow,deny
		Allow from all
	</Directory>

	ErrorLog ${APACHE_LOG_DIR}/error.log

	# Possible values include: debug, info, notice, warn, error, crit,
	# alert, emerg.
	LogLevel error

	CustomLog ${APACHE_LOG_DIR}/ssl_access.log combined

	#   SSL Engine Switch:
	#   Enable/Disable SSL for this virtual host.
	SSLEngine on

	#   A self-signed certificate
	SSLCertificateFile    /etc/ssl/certs/mydomainname.com.crt
	SSLCertificateKeyFile /etc/ssl/private/mydomainname.com.key

    # Options based on bettercrypto.org
    SSLProtocol All -SSLv2 -SSLv3
    SSLHonorCipherOrder On
    SSLCompression off
    SSLCipherSuite EDH+CAMELLIA:EDH+aRSA:EECDH+aRSA+AESGCM:EECDH+aRSA+SHA384:EECDH+aRSA+SHA256:EECDH:+CAMELLIA256:+AES256:+CAMELLIA128:+AES128:+SSLv3:!aNULL:!eNULL:!LOW:!3DES:!MD5:!EXP:!PSK:!SRP:!DSS:!RC4:!SEED:!ECDSA:CAMELLIA256-SHA:AES256-SHA:CAMELLIA128-SHA:AES128-SHA

	#   SSL Engine Options:
	#   Set various options for the SSL engine.
	#   o FakeBasicAuth:
	#     Translate the client X.509 into a Basic Authorisation.  This means that
	#     the standard Auth/DBMAuth methods can be used for access control.  The
	#     user name is the `one line' version of the client's X.509 certificate.
	#     Note that no password is obtained from the user. Every entry in the user
	#     file needs this password: `xxj31ZMTZzkVA'.
	#   o ExportCertData:
	#     This exports two additional environment variables: SSL_CLIENT_CERT and
	#     SSL_SERVER_CERT. These contain the PEM-encoded certificates of the
	#     server (always existing) and the client (only existing when client
	#     authentication is used). This can be used to import the certificates
	#     into CGI scripts.
	#   o StdEnvVars:
	#     This exports the standard SSL/TLS related `SSL_*' environment variables.
	#     Per default this exportation is switched off for performance reasons,
	#     because the extraction step is an expensive operation and is usually
	#     useless for serving static content. So one usually enables the
	#     exportation for CGI and SSI requests only.
	#   o StrictRequire:
	#     This denies access when "SSLRequireSSL" or "SSLRequire" applied even
	#     under a "Satisfy any" situation, i.e. when it applies access is denied
	#     and no other module can change it.
	#   o OptRenegotiate:
	#     This enables optimized SSL connection renegotiation handling when SSL
	#     directives are used in per-directory context.
	#SSLOptions +FakeBasicAuth +ExportCertData +StrictRequire
	<FilesMatch "\.(cgi|shtml|phtml|php)$">
		SSLOptions +StdEnvVars
	</FilesMatch>
	<Directory /usr/lib/cgi-bin>
		SSLOptions +StdEnvVars
	</Directory>

	#   SSL Protocol Adjustments:
	#   The safe and default but still SSL/TLS standard compliant shutdown
	#   approach is that mod_ssl sends the close notify alert but doesn't wait for
	#   the close notify alert from client. When you need a different shutdown
	#   approach you can use one of the following variables:
	#   o ssl-unclean-shutdown:
	#     This forces an unclean shutdown when the connection is closed, i.e. no
	#     SSL close notify alert is send or allowed to received.  This violates
	#     the SSL/TLS standard but is needed for some brain-dead browsers. Use
	#     this when you receive I/O errors because of the standard approach where
	#     mod_ssl sends the close notify alert.
	#   o ssl-accurate-shutdown:
	#     This forces an accurate shutdown when the connection is closed, i.e. a
	#     SSL close notify alert is send and mod_ssl waits for the close notify
	#     alert of the client. This is 100% SSL/TLS standard compliant, but in
	#     practice often causes hanging connections with brain-dead browsers. Use
	#     this only for browsers where you know that their SSL implementation
	#     works correctly.
	#   Notice: Most problems of broken clients are also related to the HTTP
	#   keep-alive facility, so you usually additionally want to disable
	#   keep-alive for those clients, too. Use variable "nokeepalive" for this.
	#   Similarly, one has to force some clients to use HTTP/1.0 to workaround
	#   their broken HTTP/1.1 implementation. Use variables "downgrade-1.0" and
	#   "force-response-1.0" for this.
	BrowserMatch "MSIE [2-6]" \
		nokeepalive ssl-unclean-shutdown \
		downgrade-1.0 force-response-1.0
	# MSIE 7 and newer should be able to use keepalive
	BrowserMatch "MSIE [17-9]" ssl-unclean-shutdown

</VirtualHost>
</IfModule>
#+END_SRC

Enable your site with:

#+BEGIN_SRC: bash
a2ensite
#+END_SRC

then select the domain name and reload.

#+BEGIN_SRC: bash
service apache2 reload
#+END_SRC

and alter permissions:

#+BEGIN_SRC: bash
chmod -R 755 /var/www/$HOSTNAME/htdocs
chown -R www-data:www-data /var/www/$HOSTNAME/htdocs
#+END_SRC

Open a browser and visit http://$HOSTNAME/install.php, then fill out the details.  Once everything has been accepted without errors:

#+BEGIN_SRC: bash
rm /var/www/$HOSTNAME/htdocs/install.php
#+END_SRC

Add a few extra mime types:

#+BEGIN_SRC: bash
emacs /var/www/$HOSTNAME/htdocs/conf/mime.conf
#+END_SRC

Append the following:

#+BEGIN_SRC: bash
ogv     video/ogg
mp4     video/mp4
webm    video/webm
#+END_SRC

Save and exit.

If you need to be able to upload large files to the wiki then edit */etc/php5/apache2/php.ini* and set *upload_max_filesize* accordingly. If the directory */etc/php5/apache2* doesn't exist then you will need to install the package *libapache2-mod-php5*.

Now you can visit your wiki and begin editing.

** Install Bitmessage

#+BEGIN_VERSE
/The weakness of mass surveillance is that it can very easily be made much more expensive through changes in technical standards: pervasive, end-to-end encryption can quickly make indiscriminate surveillance impossible on a cost-effective basis/

-- Edward J. Snowden, testimony to the EU parliament
#+END_VERSE

*** A new kind of Email
[[https://bitmessage.org][Bitmessage]] is a new type of messaging system intended to fulfill the same role as email, but without the security problems.  In particular, Bitmessage attempts to not just encrypt the content but also the metadata.  It's message broadcasting system makes it exceedingly difficult for an attacker to know which computer a message is destined for.  The only way you know whether a message has been sent to you is whether you are able to decrypt it from the passing stream of messages.

Although similar to Bitcoin in some regards, such as "/proof of work/", Bitmessage has no block chain and messages are only buffered for approximately three days after which they are deleted from any given node.

Installing Bitmessage as a daemon will increase the size of the network, and therefore the level of security for all users.

*** The Daemon

Install from the current source code.

#+BEGIN_SRC: bash
apt-get install python screen
cd /tmp
git clone https://github.com/Bitmessage/PyBitmessage.git
cd PyBitmessage
make install
#+END_SRC

Now create the daemon.

#+BEGIN_SRC: bash
emacs /etc/init.d/pybitmessage
#+END_SRC

Add the following text:

#+BEGIN_SRC: bash
#!/bin/bash
# /etc/init.d/bitmessage

### BEGIN INIT INFO
# Provides:          pybitmessage
# Required-Start:    $remote_fs $syslog
# Required-Stop:     $remote_fs $syslog
# Default-Start:     2 3 4 5
# Default-Stop:      0 1 6
# Short-Description: starts bitmessage as a background daemon, suitable for servers
# Description:       This file should be used to construct scripts to be
#                    placed in /etc/init.d.
### END INIT INFO

# Author: Super-Nathan <BM-Gu2k3Wy2hpTMYBxSoM2937SPcuU6xzEj>

#Settings
SERVICE='pybitmessage'
LOGFILE='/dev/null' # this disables logging
# LOGFILE='/var/log/bitmessage.log' # comment out the above line and un-comment this line to save a log
COMMAND="python bitmessagemain.py > $LOGFILE"
USERNAME='bitmsg'
NICELEVEL=19 # from 0-19 the bigger the number, the less the impact on system resources
HISTORY=1024
PBM_LOCATION="/usr/local/share/pybitmessage"
INVOCATION="nice -n ${NICELEVEL} ${COMMAND}"
PATH='/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/bin/core_perl:/sbin:/usr/sbin:/bin:/usr/local/share/pybitmessage'




bm_start() {
echo "Starting $SERVICE..."
cd ${PBM_LOCATION}
su --command "screen -h ${HISTORY} -dmS ${SERVICE} ${INVOCATION}" $USERNAME
}


bm_stop() {
echo "Stopping $SERVICE"
su --command "screen -p 0 -S ${SERVICE} -X stuff "'^C'"" $USERNAME
}


#Start-Stop here
case "$1" in
  start)
    bm_start
    ;;
  stop)
    bm_stop
    ;;
  restart)
    bm_stop
    sleep 60s
    bm_start
    ;;
    *)
  echo "Usage: $0 {start|stop|restart}"
  exit 1
  ;;
esac

exit 0
#+END_SRC

Save and exit.

Add a user which will be specifically for Bitmessage.  Since bitmessage is still a relatively young and experimental project, this adds further compartmentalisation such that if there are any bugs within PyBitmessage then an attacker can't neccessarily gain control of root or any other user account.  Here we create a user called /bitmsg/ and give it a long random password.

#+BEGIN_SRC: bash
adduser bitmsg
#+END_SRC

Create a /keys.dat/ file which is used to configure Bitmessage.

#+BEGIN_SRC: bash
mkdir /home/bitmsg/.config
mkdir /home/bitmsg/.config/PyBitmessage
emacs /home/bitmsg/.config/PyBitmessage/keys.dat
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
[bitmessagesettings]
settingsversion = 7
port = 8444
timeformat = %%a, %%d %%b %%Y  %%I:%%M %%p
blackwhitelist = black
startonlogon = false
minimizetotray = false
showtraynotifications = false
startintray = false
socksproxytype = none
sockshostname = localhost
socksport = 9050
socksauthentication = false
sockslisten = false
socksusername =
sockspassword =
keysencrypted = false
messagesencrypted = false
defaultnoncetrialsperbyte = 640
defaultpayloadlengthextrabytes = 14000
minimizeonclose = false
maxacceptablenoncetrialsperbyte = 0
maxacceptablepayloadlengthextrabytes = 0
userlocale = system
namecoinrpctype = namecoind
namecoinrpchost = localhost
namecoinrpcuser =
namecoinrpcpassword =
namecoinrpcport = 8336
sendoutgoingconnections = True
daemon = true
#+END_SRC

Save and exit. Then enable the daemon and run it.

#+BEGIN_SRC: bash
rm -f /tmp/-usr-local-share-pybitmessage-*.lock
chown -R bitmsg:bitmsg /home/bitmsg
chmod +x /etc/init.d/pybitmessage
update-rc.d pybitmessage defaults
service pybitmessage start
#+END_SRC

Now open port 8444 on your internet router or firewall and direct it to the BBB.

*** Using Bitmessage
Although in principle it would be possible to send Bitmessages directly from the BBB, in practice the /proof of work/ requirement would mean that it would take an infeasibly long time to send messages, and the computational workload would likely greatly impair the performance of other services also running on the system.  So to send and receive Bitmessages it's better to just install the client on a laptop or desktop machine.

The easiest way to install the client is either to download it from [[https://bitmessage.org][bitmessage.org]] or to get the latest build from Github as follows:

#+BEGIN_SRC: bash
cd /tmp
git clone https://github.com/Bitmessage/PyBitmessage.git
cd PyBitmessage
make install
pybitmessage
#+END_SRC

*** Connect to Email

TODO: how to connect Bitmessage to an email client.

#+BEGIN_SRC: bash
cd /tmp
wget http://freedombone.uk.to/notbit.tar.gz
#+END_SRC

Verify it.

#+BEGIN_SRC: bash
sha256sum notbit.tar.gz
972fdc9cbb8034141282337dcd5e557bce57969ff6bd1d607da89bd93cc7bb68
#+END_SRC

Extract and install it.

#+BEGIN_SRC: bash
tar -xzvf notbit.tar.gz
cd notbit
apt-get install dh-autoreconf
./autogen.sh --prefix=/home/myusername
make
make install
#+END_SRC

** Overcome restrictive environments

#+BEGIN_VERSE
/Censorship reflects a society's lack of confidence in itself. It is a hallmark of an authoritarian regime./

-- Potter Stewart
#+END_VERSE

In some environments, such as behind corporate firewalls or under regimes hostile towards the idea of open access to knowledge and information you may find that you're not able to use tools such as /ssh/ to get access to the BBB.  In the worst case all ports other than 80 and 443 may be blocked.

In that scenario you can use a tool called [[http://code.google.com/p/shellinabox/][shellinabox]] to log into your BBB via your web site rather than via a terminal.  This means that you can administrate your system from any device which has a web browser and keyboard.

#+BEGIN_SRC: bash
apt-get install shellinabox libapache2-mod-proxy-html
#+END_SRC

Update your Apache configuration.

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
emacs /etc/apache2/sites-available/$HOSTNAME
#+END_SRC

Within the section which begins with *<VirtualHost *:443>* add the following, replacing /mydomainname.com/ with your domain name and /myusername/ with your username.

#+BEGIN_SRC: bash
  <Location /shell>
    ProxyPass http://localhost:4200/
    Order allow,deny
    Allow from all

    AuthName "Authentication for shellinabox"
    AuthUserFile /home/mydomainname.com/public_html/.htpasswd
    AuthGroupFile /home/mydomainname.com/public_html/.htgroup
    AuthType Basic
    Require group shellinabox
    Require user myusername
  </Location>
#+END_SRC

Save and exit, then create a login password.  It's recommended that the password be a long random string and that you then access it using a password manager such as KeepassX.

#+BEGIN_SRC: bash
mkdir /home/$HOSTNAME
mkdir /home/$HOSTNAME/public_html
htpasswd -c /home/$HOSTNAME/public_html/.htpasswd myusername
#+END_SRC

Create a user group.

#+BEGIN_SRC: bash
emacs /home/$HOSTNAME/public_html/.htgroup
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
shellinabox: myusername
#+END_SRC

Save and exit, then restart Apache.

#+BEGIN_SRC: bash
a2enmod proxy_http
service apache2 restart
#+END_SRC

Now with a web browser navigate to https://mydomainname.com/shell and log in.

If you're in a very locked down environment where access to web sites is severely restricted then as a last resort you may be able to use a command line browser, such as [[https://en.wikipedia.org/wiki/Lynx_%28web_browser%29][lynx]] from within /shellinabox/.

** Set up a mailing list

#+BEGIN_VERSE
/All over the world there are many people who are united in creating software, content, and culture that is freely available for others to share, enjoy and enrich their lives. Together we believe that freedom is good. We believe it helps people do good things, make better choices, and lead safer and more secure lives. Together we are a community united by this belief./

-- Jono Bacon
#+END_VERSE

*** Public mailing list
Email mailing lists are old skool but still remain as a common and easy way of communicating on the internet.  If you're running a public organisation such as an open source project or community group then you may want to set one up.

**** Installation

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
apt-get install mailman
newlist mailman
#+END_SRC

Enter an email address for the list administrator and a password.

#+BEGIN_SRC: bash
emacs /etc/mailman/mm_cfg.py
#+END_SRC

Set *MTA=None* and change *http:* to *https:*, then save and exit.

Add some settings.

#+BEGIN_SRC: bash
emacs /etc/exim4/conf.d/main/04_mailman_options
#+END_SRC

Add the following, replacing /mydomainname.com/ with your domain name.

#+BEGIN_SRC: bash
# Mailman macro definitions

# Home dir for the Mailman installation
MM_HOME=/var/lib/mailman

# User and group for Mailman
MM_UID=list
MM_GID=list

#
# Domains that your lists are in - colon separated list
# you may wish to add these into local_domains as well
domainlist mm_domains=mydomainname.com

# The path of the Mailman mail wrapper script
MM_WRAP=MM_HOME/mail/mailman
#
# The path of the list config file (used as a required file when
# verifying list addresses)
MM_LISTCHK=MM_HOME/lists/${lc::$local_part}/config.pck
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/exim4/conf.d/main/000_localmacros
#+END_SRC

Append the following:

#+BEGIN_SRC: bash
SYSTEM_ALIASES_PIPE_TRANSPORT = address_pipe
SYSTEM_ALIASES_USER = list
SYSTEM_ALIASES_GROUP = list
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/exim4/conf.d/acl/30_exim4-config_check_rcpt
#+END_SRC

Append the following, before the final /accept/:

#+BEGIN_SRC: bash
  # Do callback verification unless Mailman incoming bounce
  deny !local_parts = *-bounces : *-bounces+*
     !verify = sender/callout=30s,defer_ok
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs
/etc/exim4/conf.d/router/450_exim4-config_mailman_aliases
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
mailman:
  driver = accept
  domains = +mm_domains
  require_files = MM_LISTCHK
  local_part_suffix_optional
  local_part_suffix = -admin : \
    -bounces   : -bounces+*  : \
    -confirm   : -confirm+*  : \
    -join      : -leave      : \
    -owner     : -request    : \
    -subscribe : -unsubscribe
  transport = mailman_transport
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/exim4/conf.d/transport/40_exim4-config_mailman_pipe
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
mailman_transport:
  driver = pipe
  command = MM_WRAP \
    '${if def:local_part_suffix \
    {${sg{$local_part_suffix}{-(\\w+)(\\+.*)?}{\$1}}} \
    {post}}' \
    $local_part
  current_directory = MM_HOME
  home_directory = MM_HOME
  user = MM_UID
  group = MM_GID
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chown root:list /var/lib/mailman/mail/mailman
update-exim4.conf.template -r
update-exim4.conf
service exim4 restart
emacs /etc/apache2/conf.d/mailman
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
Alias /pipermail /var/lib/mailman/archives/public
Alias /images/mailman /usr/share/images/mailman
<directory /var/lib/mailman/archives/public>
    DirectoryIndex index.html
</directory>
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/$HOSTNAME
#+END_SRC

Add the following to the 443 section.

#+BEGIN_SRC: bash
  <Location /mailman>
    Options Indexes FollowSymLinks MultiViews
    Order allow,deny
    Allow from all

    RedirectMatch ^/$ /cgi-bin/mailman/listinfo
  </Location>
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
service apache2 restart
#+END_SRC

Now add your mailing list.  The list name should not include any spaces.

#+BEGIN_SRC: bash
newlist mymailinglistname
#+END_SRC

With a browser visit https://$HOSTNAME/cgi-bin/mailman/admin/mymailinglistname to configure the mailing list.

Under *General Options* add an email address for a moderator (could be the same as the administrator) and click *Submit your changes*.

Under *Privacy Options* set steps required for subscription to *Confirm and approve* and click *Submit your changes*.

Also change these settings for the account within https://$HOSTNAME/cgi-bin/mailman/admin/mailman

Then to test that the mailing list works:

#+BEGIN_SRC: bash
exim -d+route -bt mymailinglistname@$HOSTNAME
#+END_SRC

If everything is working then this shouldn't show any problems.

**** Using the mailing list
Direct subscribers towards:

#+BEGIN_SRC: bash
https://mydomainname.com/cgi-bin/mailman/listinfo/mymailinglistname
#+END_SRC

To administrate the list visit:

#+BEGIN_SRC: bash
https://mydomainname.com/cgi-bin/mailman/admin/mymailinglistname
#+END_SRC

To add another mailing list:

#+BEGIN_SRC: bash
newlist mymailinglistname
#+END_SRC

To delete a mailing list:

#+BEGIN_SRC: bash
rmlist -a mymailinglistname
#+END_SRC

*** Private (encrypted) mailing list
In addition to conventional public email lists it's also possible to set up a private mailing list which is only readable by members.  A private email list uses [[https://en.wikipedia.org/wiki/GNU_Privacy_Guard][GPG]] and a public/private key pair for the server which can then be used to send emails to the list in an encrypted form.  The email addresses and public GPG keys of members may be added to the list so that any new messages can be distributed to them in a secure manner.

Private mailing lists are likely to be able to keep the contents of the discussion out of the clutches of warrantless mass surveillance but, as with all conventional email, it won't prevent such systems from generating social graphs of who is communicating with the list since the /from/ and /to/ attributes are always transmitted in the clear.

**** Installation
#+BEGIN_SRC: bash
apt-get install schleuder
#+END_SRC

Edit the configuration:

#+BEGIN_SRC: bash
emacs /etc/schleuder/schleuder.conf
#+END_SRC

Set the following parameters, replacing /mydomainname.com/ with your domain name:

#+BEGIN_SRC: bash
smtp_port: 465
superadminaddr: root@mydomainname.com
#+END_SRC

Save and exit.

Get your GPG public key, replacing /myGPGkeyID/ with your GPG key ID:

#+BEGIN_SRC: bash
export MYKEYID=myGPGkeyID
gpg --search-keys $MYKEYID
gpg --output /tmp/mypublickey.txt --armor --export $MYKEYID
#+END_SRC

Then to create a mailing list, replacing /mydomainname.com/ with your domain name, /myusername/ with your username and /mailinglistname/ with the name of the mailing list.  /mailinglistname/ should be all one word, with no spaces.

#+BEGIN_SRC: bash
export MAILINGLISTNAME=mailinglistname
export MYUSERNAME=myusername
export HOSTNAME=mydomainname.com
export EMAILADDRESS=$MYUSERNAME@$HOSTNAME
schleuder-newlist $MAILINGLISTNAME@$HOSTNAME -realname "mailing list name" -adminaddress $EMAILADDRESS -initmember $EMAILADDRESS -initmemberkey /tmp/mypublickey.txt -nointeractive
#+END_SRC

Now add a mailing list rule:

#+BEGIN_SRC: bash
emailrule $MYUSERNAME $MAILINGLISTNAME@$HOSTNAME $MAILINGLISTNAME
#+END_SRC

Edit your Mutt configuration.

#+BEGIN_SRC: bash
emacs /home/$MYUSERNAME/.muttrc
#+END_SRC

Search for the /mailboxes/ parameter and add "=mailinglistname".  For example:

#+BEGIN_SRC: bash
mailboxes = =Sent =Drafts =mailinglistname
#+END_SRC

Save and exit.

Update Exim routing.

#+BEGIN_SRC: bash
emacs /etc/exim4/conf.d/router/550_exim4-config_schleuder
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
schleuder:
  debug_print = "R: schleuder for $local_part@$domain"
  driver = accept
  local_part_suffix_optional
  local_part_suffix = +* : -bounce : -sendkey
  domains = +local_domains
  user = schleuder
  group = schleuder
  require_files = schleuder:+/var/lib/schleuder/$domain/${local_part}
  transport = schleuder_transport
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
emacs /etc/exim4/conf.d/transport/30_exim4-config_schleuder
#+END_SRC

Add the following.

#+BEGIN_SRC: bash
schleuder_transport:
  debug_print = "T: schleuder_transport for $local_part@$domain"
  driver = pipe
  home_directory = "/var/lib/schleuder/$domain/$local_part"
  command = "/usr/bin/schleuder $local_part@$domain"
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chown -R schleuder:schleuder /var/lib/schleuder
update-exim4.conf.template -r
update-exim4.conf
service exim4 restart
useradd -d /var/schleuderlists -s /bin/false schleuder
adduser Debian-exim schleuder
usermod -a -G mail schleuder
#+END_SRC

Test the routing.

#+BEGIN_SRC: bash
exim -d -bt mailinglistname@mydomainname.com
#+END_SRC

**** Importing the public key of the mailing list
Before you can use the mailing list you will first need to import its public key.  How you do this depends upon which email client you're using.

***** Using Mutt
Send an email to /mailinglistname-sendkey@mydomainname.com/ to have the list public key emailed to you.

When you receive the email open it and press *CTRL-k* to import it.
***** Using Thunderbird
Send an email to /mailinglistname-sendkey@mydomainname.com/ to have the list public key emailed to you.

When you receive the email open it, select all the text with *CTRL-a* then *CTRL-c*.

On the menu select *OpenPGP* followed by *Key Management*.

You will now see a new menu bar.  Select *Edit* followed by *Import keys from clipboard*.

Click on *Import* followed by *Ok*.
**** Using the list
To obtain the public keys of list members send an email to /mailinglistname-request@mydomainname.com/ containing *X-LIST-KEYS* in the message body.

To add a member: *X-ADD-MEMBER: othermember@otherdomain.net*

An example of adding a public key to the list:

#+BEGIN_SRC: bash
X-ADD-KEY:
-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: GnuPG v1.4.9 (GNU/Linux)

mQGiBEjVO7oRBADQvT6wtD2IzzIiK0NbrcilCKCp4MWb8cYXTXguwPQI6y0Nerz4
dsK6J0X1Vgeo02tqA4xd3EDK8rdqL2yZfl/2egH8+85R3gDk+kqkfEp4pwCgp6VO
[...]
pNlF/qkaWwRb048h+iMrW21EkouLKTDPFkdFbapV2X5KJZIcfhO1zEbwc1ZKF3Ju
Q9X5GRmY62hz9SCZnsC0jeYAni8OUQV9NXfXlS/vePBUnOL08NQB
=xTv3
-----END PGP PUBLIC KEY BLOCK-----
#+END_SRC

To get details for a member: *X-GET-MEMBER: othermember@otherdomain.net*

To delete a member: *X-DELETE-MEMBER: othermember@otherdomain.net*

To delete a public key: *X-DELETE-KEY: keyID*

You can unsubscribe from the list with *X-UNSUBSCRIBE* in the message body.

*** Decentralised mailing list
A disadvantage with encrypted mailing lists which use the conventional email system is that there is a single server on which the list resides, and this creates a single point of failure and a bandwidth bottleneck for more heavily subscribed lists.  If the mailing list server goes down for whatever reason then that may cause a lot of disruption to its users.

An alternative is to use a decentralised mailing list, implemented using Bitmessage.  On your local machine (not the BBB) you can make a private mailing list which is difficult to censor and where there is no single point of failure.  This type of mailing list is known as a "/chan/".

With Bitmessage if any one computer goes offline then the conversation can still keep going since there is no central mailing list server. Bitmessages are also encrypted with public/private key pairs and the manner in which the system operates makes it very difficult for the surveillance apparatus to exfiltrate the social graph of list users.

On a Debian based system:

#+BEGIN_SRC: bash
sudo apt-get install makepasswd
#+END_SRC

or on an RPM based system:

#+BEGIN_SRC: bash
sudo yum install makepasswd
#+END_SRC

Create a name for your mailing list.  This will be a random string.

#+BEGIN_SRC: bash
makepasswd -c 40
#+END_SRC

Keep a note of this.

Run the Bitmessage client and on the menu select *File/Join-Create Chan/Create new chan*

Enter the random string which you created as the name of the mailing list.  Also take a note of the BM address which is created.

You can hand out the random string used to generate the mailing list and its corresponding BM address to fellow members, either within a bitmessage or on paper or via [[https://en.wikipedia.org/wiki/Sneakernet][sneakernet]] or in a GPG/PGP encrypted email or via an XMPP+OTR or Friendica private message.  Once others have those two pieces of data then they will be able to join.

To make the list easier to identify, rather than just appearing as a random string, then under the *Your Identities* tab right click on it and select *Set Avatar* and assign a suitable icon.

The disadvantage of this type of mailing list is that it's not possible for any one participant to act as a list moderator, or in other words each participant must do their own moderation.  That's ok if the size of the group is small, but if it's larger then anyone spamming or trolling the list can make things miserable for the others.
** Add a Convergenge notary
Convergence is a secure replacement for the Certificate Authority System. Rather than employing a traditionally hard-coded list of immutable and largely untrusted CAs, Convergence allows you to configure a dynamic set of Notaries which use network perspective to validate your communication.  For more details see [[http://convergence.io][convergence.io]] or [[http://www.youtube.com/watch?v=Z7Wl2FW2TcA][this talk which explains the concepts]].

*** Installation

#+BEGIN_SRC: bash
apt-get install python python-twisted-web python-twisted-names python-m2crypto python-openssl
cd /tmp
git clone https://github.com/fuzzgun/convergence
cd convergence/server
python ./setup.py install
#+END_SRC

Generate a key pair:

#+BEGIN_SRC: bash
convergence gencert
#+END_SRC

When asked for a challenge password just hit *Enter* a couple of times. Then move the key pair to the appropriate directories as follows.

#+BEGIN_SRC: bash
mv mynotary.key /etc/ssl/private
chmod 400 /etc/ssl/private/mynotary.key
mv mynotary.pem /etc/ssl/certs
#+END_SRC

Now create the database:

#+BEGIN_SRC: bash
rm /var/lib/convergence/convergence.db
convergence createdb
#+END_SRC

Create an initialisation script:

#+BEGIN_SRC: bash
emacs /etc/init.d/convergence
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
adduser converg
#+END_SRC

The details for the user don't especially matter, but give them a long random password.

#+BEGIN_SRC: bash
chown -R converg:converg /home/converg
chmod +x /etc/init.d/convergence
update-rc.d convergence defaults
service convergence start
#+END_SRC

Generate a notary bundle:

#+BEGIN_SRC: bash
convergence bundle
#+END_SRC

Enter your name, nickname, handle or whatever.

For the bundle location enter https://mydomainname.com/convergence.notary

For the Hostname enter your domain name

For SSL port enter *8433* and for HTTP port nter *8432*

For the pem file enter */etc/ssl/certs/mynotary.pem*

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
mv mynotarybundle.notary /var/www/$HOSTNAME/htdocs/convergence.notary
chown www-data:www-data /var/www/$HOSTNAME/htdocs/convergence.notary
#+END_SRC

Now open ports 8432 and 8433 on your internet router or firewall and direct it to the BBB.

*** Using Convergence
On a computer which is not the BBB (your laptop, etc):

Install the browser plugin by navigating to https://addons.mozilla.org/en-us/firefox/addon/convergence-extra/

After installation restart your browser.

You will notice that an icon appears in the top right corner of the browser, which resembles a lock and two plus signs.  Click on the down arrow to the right of it and select *options*.
** Install a microblog

#+BEGIN_VERSE
/If you want to have more control over how you interact on the web, and regain your freedom, privacy and autonomy from outside interference, you need to start moving towards using programs like GNU Social/

-- Jason Self
#+END_VERSE

For a microblog you will need a separate domain/subdomain, so see [[Setting up a web site]] for details of how to create an Apache configuration for your microblog.  If you're using freedns then you will need to create a new subdomain.

Install some dependencies:

#+BEGIN_SRC: bash
apt-get install php5-xcache php-gettext php5-curl php5-gd php5-mysql
#+END_SRC

Download GNU Social

#+BEGIN_SRC: bash
cd /tmp
wget http://freedombone.uk.to/gnu-social.tar.gz
#+END_SRC

Verify it.

#+BEGIN_SRC: bash
sha256sum gnu-social.tar.gz
1f886241c7f1a175e7be3cccbcb944ab6c03617fb75aefa4d62d37abed87d2b4
#+END_SRC

Extract the files and set permissions on them, where /mydomainname.com/ is your domain name.

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
tar zxf gnu-social.tar.gz
rm -rf /var/www/$HOSTNAME/htdocs
mv statusnet-gnu-social /var/www/$HOSTNAME/htdocs
chmod a+w /var/www/$HOSTNAME/htdocs
chown www-data:www-data /var/www/$HOSTNAME/htdocs
chmod a+w /var/www/$HOSTNAME/htdocs/avatar
chmod a+w /var/www/$HOSTNAME/htdocs/background
chmod a+w /var/www/$HOSTNAME/htdocs/file
chmod +x /var/www/$HOSTNAME/htdocs/scripts/maildaemon.php
#+END_SRC

Edit the Apache access settings.

#+BEGIN_SRC: bash
emacs /var/www/$HOSTNAME/htdocs/.htaccess
#+END_SRC

Add the following:

#+BEGIN_SRC: bash
<IfModule mod_rewrite.c>
  RewriteEngine On
  RewriteBase /

  ## Uncomment these if having trouble with API authentication
  ## when PHP is running in CGI or FastCGI mode.
  #
  #RewriteCond %{HTTP:Authorization} ^(.*)
  #RewriteRule ^(.*) - [E=HTTP_AUTHORIZATION:%1]

  RewriteCond %{REQUEST_FILENAME} !-f
  RewriteCond %{REQUEST_FILENAME} !-d
  RewriteRule (.*) index.php?p=$1 [L,QSA]
</IfModule>

<FilesMatch "\.(ini)">
  Order allow,deny
</FilesMatch>
#+END_SRC

Save and exit, then create a database.

#+BEGIN_SRC: bash
mysql -u root -p
create database gnusocial;
CREATE USER 'gnusocialadmin'@'localhost' IDENTIFIED BY 'gnusocialpassword';
GRANT ALL PRIVILEGES ON gnusocial.* TO 'gnusocialadmin'@'localhost';
quit
#+END_SRC

Add the mailer script to the aliases file:

#+BEGIN_SRC: bash
emacs /etc/aliases
#+END_SRC

Add the following, replacing /mydomainname.com/ with your domain name.

#+BEGIN_SRC: bash
www-data: root
*: /var/www/mydomainname.com/htdocs/scripts/maildaemon.php
#+END_SRC

Save and exit.  Update the aliases by typing:

#+BEGIN_SRC: bash
newaliases
#+END_SRC

Then with a web browser navigate to:

https://$HOSTNAME/install.php

Set a name for the site.

Server SSL: enable

Hostname: localhost

Type: MySql

Name: gnusocial

DB username: gnusocialadmin

DB Password; your gnu social admin password goes here

Administrator nickname: myusername

Administrator password: mylongrandompassword

Subscribe to announcements: ticked

Site profile: Community

Press the *Submit* button.  It may take a few minutes, so don't be concerned that it has crashed.  When the process completes you will see a lot of "Strict standards" warnings which you can ignore.

Navigate to http://$HOSTNAME/gnusocial and you can then complete the configuration via the *Admin* section on the header bar.  Some recommended admin settings are:

Under the *Site* settings:

    Text limit: 140

    Dupe Limit: 60000

Under the *User* settings:

    Bio limit: 1000

Under the *Access* settings:

    /Invite only/ ticked

Under the License section select a license if you wish.  Details for Creative Commons licenses [[https://creativecommons.org/licenses/][can be found here]].  If you only intend to do private microblogging then just leave these settings as they are.

If you want to invite more users then click on the big button *Invite more colleagues*, then enter their email addresses and hit the *send* button.  The invite only configuration which you've just installed is useful because it prevents spammers, or other [[https://en.wikipedia.org/wiki/Joint_Threat_Research_Intelligence_Group]["bad actors"]], from clogging your system with nonsense.

Edit the config file.

#+BEGIN_SRC: bash
emacs /var/www/$HOSTNAME/htdocs/config.php
#+END_SRC

Change the ssl setting from *always* to *sometimes*, hten save and exit.

So, you're now microblogging on the open web, with no companies in the middle.  Congratulations!  To find some other people to connect to you can try searching other nodes listed at http://gnu.io/try/

When following other GNU Social users enter the URL of your profile. For example, https://mygnusocialdomain/myusername

** Install Mediagoblin

#+BEGIN_VERSE
/The silos that are the main current points of media sharing are not only vulnerable to attacks on free speech, but also hamper important grassroots economic activity by privileging the interests of a tiny minority over those of most of the world./
#+END_VERSE

Mediagoblin allows you to have a YouTube/Soundcloud/Flickr/Picasa type of site to share your pictures, videos or audio files.  An advantage of not having any company in the middle is that you can't be arbitrarily censored without any explanation, as seems to frequently occur on YouTube.  It is recommended that you use media formats which are not encumbered by patents, such as /ogg/ or /ogv/.

For a mediagoblin site it is recommended to use a separate domain/subdomain, so see [[Setting up a web site]] for details of how to create an Apache configuration for your microblog.  If you're using freedns then you will need to create a new subdomain.

Install some dependencies.

#+BEGIN_SRC: bash
apt-get install git-core python python-dev python-lxml python-imaging python-virtualenv python-gst0.10 libjpeg8-dev sqlite3 libapache2-mod-fcgid gstreamer0.10-plugins-base gstreamer0.10-plugins-bad gstreamer0.10-plugins-good gstreamer0.10-plugins-ugly gstreamer0.10-ffmpeg python-numpy python-scipy libsndfile1-dev
#+END_SRC

Create a user, replacing /mymediagoblindomain/ with the domain name for your mediagoblin site.

#+BEGIN_SRC: bash
export HOSTNAME=mymediagoblindomain
adduser mediagoblin
#+END_SRC

Give the user a long random password.

#+BEGIN_SRC: bash
mkdir -p /srv/$HOSTNAME
chown -hR mediagoblin:mediagoblin /srv/$HOSTNAME
su - mediagoblin
export HOSTNAME=mymediagoblindomain
cd /srv/$HOSTNAME
git clone git://gitorious.org/mediagoblin/mediagoblin.git
cd mediagoblin
git submodule init
git submodule update
virtualenv --system-site-packages .
./bin/python setup.py develop
./bin/easy_install flup
cp mediagoblin.ini mediagoblin_local.ini
cp paste.ini paste_local.ini
emacs mediagoblin_local.ini
#+END_SRC

Change *email_sender_address* to your email address and set *email_debug_mode* to false.  Also append the following to the bottom of the file, under the *plugins* section.

#+BEGIN_SRC: bash
[[mediagoblin.media_types.audio]]
[[mediagoblin.media_types.video]]
[[mediagoblin.media_types.stl]]
#+END_SRC

Then save and exit.

#+BEGIN_SRC: bash
./bin/pip install scikits.audiolab
./bin/gmg dbupdate
exit # to go back to the root user
emacs /etc/init.d/mediagoblin
#+END_SRC

Add the following, replacing /mymediagoblindomain/ with the domain name for your mediagoblin site.

#+BEGIN_SRC: bash
#!/bin/bash
# /etc/init.d/mediagoblin

### BEGIN INIT INFO
# Provides:          mediagoblin
# Required-Start:    $remote_fs $syslog
# Required-Stop:     $remote_fs $syslog
# Default-Start:     2 3 4 5
# Default-Stop:      0 1 6
# Short-Description: starts mediagoblin
# Description:       Other methods may work, but I found this the easiest
### END INIT INFO

# Author: Bob Mottram <bob@robotics.uk.to>

#Settings
SERVICE='mediagoblin'
LOGFILE='/srv/mymediagoblindomain/mediagoblin.log'
COMMAND="./lazyserver.sh > $LOGFILE"
USERNAME='mediagoblin'
NICELEVEL=15 # from 0-19 the bigger the number, the less the impact on system resources
HISTORY=1024
MG_LOCATION="/srv/mymediagoblindomain/mediagoblin"
INVOCATION="nice -n ${NICELEVEL} ${COMMAND}"
PATH='/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/bin/core_perl:/sbin:/usr/sbin:/bin'



mg_start() {
echo "Starting $SERVICE..."
cd ${MG_LOCATION}
su --command "screen -h ${HISTORY} -dmS ${SERVICE} ${INVOCATION}" $USERNAME
}


mg_stop() {
echo "Stopping $SERVICE"
su --command "screen -p 0 -S ${SERVICE} -X stuff "'^C'"" $USERNAME
}


#Start-Stop here
case "$1" in
  start)
    mg_start
    ;;
  stop)
    mg_stop
    ;;
  restart)
    mg_stop
    sleep 10s
    mg_start
    ;;
    *)
  echo "Usage: $0 {start|stop|restart}"
  exit 1
  ;;
esac

exit 0
#+END_SRC

Save and exit.

#+BEGIN_SRC: bash
chmod +x /etc/init.d/mediagoblin
update-rc.d mediagoblin defaults
service mediagoblin start
#+END_SRC

Edit the Apache configuration for your mediagoblin site.

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/mymediagoblindomain
#+END_SRC

Delete the existing configuration (in Emacs it's CTRL-x h then CTRL-w) and paste the following, replacing /mymediagoblindomain/ with your mediagoblin domain name and /myusername@mydomainname.com/ with your email address.

#+BEGIN_SRC: bash
<VirtualHost *:80>
    ServerAdmin myusername@mydomainname.com

    DocumentRoot /srv/mymediagoblindomain/mediagoblin
    ServerName mymediagoblindomain

    <Directory />
        Options FollowSymLinks
        AllowOverride None
    </Directory>
    <Directory /srv/mymediagoblindomain/mediagoblin/>
        Options Indexes FollowSymLinks MultiViews
        AllowOverride All
        Order allow,deny
        allow from all
    </Directory>

    LogLevel warn

    ProxyVia On

    ProxyRequests off
    ProxyPreserveHost on

    ProxyPass / http://localhost:6543/

    ErrorLog "/var/log/apache2/error_log"
    CustomLog "/var/log/apache2/access_log" combined

    RewriteEngine On
    RewriteOptions Inherit
</VirtualHost>
#+END_SRC

Save and exit.

Now in a browser visit http://mymediagoblindomain and create a user.  If you wish this to be a single user installation to prevent a lot of spammers signing up.

#+BEGIN_SRC: bash
emacs /srv/mymediagoblindomain/mediagoblin/mediagoblin_local.ini
#+END_SRC

Then set:

#+BEGIN_SRC: bash
allow_registration = false
#+END_SRC

Save and exit.

** Install Tripwire

#+BEGIN_VERSE
/...by the time you get done with all of that, we have a freedom box/

-- Eben Moglen
#+END_VERSE

Tripwire will try to detect any intrusions into your system. It's a good idea to install it after you have installed all of the other programs which you intend to use.

#+BEGIN_SRC: bash
apt-get install tripwire
export HOSTNAME=mydomainname.com
cd /etc/tripwire
cp arm-local.key $HOSTNAME-local.key
cp site.key $HOSTNAME-site.key
tripwire --init
tripwire --update-policy --secure-mode low /etc/tripwire/twpol.txt
tripwire --check --interactive
#+END_SRC

you will be asked for two passphrases ("site" and "local").  Make a note of these.

Turn off reporting of changes to system logs.

#+BEGIN_SRC: bash
emacs /etc/tripwire/twcfg.txt
#+END_SRC

Set *SYSLOGREPORTING* to false and comment out the line, then save and exit.

#+BEGIN_SRC: bash
emacs /etc/tripwire/twpol.txt
#+END_SRC

Comment out the line:

#+BEGIN_SRC: bash
	/var/log		-> $(SEC_CONFIG) ;
#+END_SRC

Then save and exit.

If you subsequently install any more packages or make configuration changes then update the policy again with:

#+BEGIN_SRC: bash
tripwire --update-policy --secure-mode low /etc/tripwire/twpol.txt
#+END_SRC

Also, to look for any rootkits.

#+BEGIN_SRC: bash
apt-get install rkhunter
#+END_SRC

* Router/Firewall ports
The following ports on your internet router/firewall should be forwarded to the BBB.

| Protocol      |     Port/s |
|---------------+------------|
| Gopher        |         70 |
| HTTP          |         80 |
| HTTPS         |        443 |
| IMAP          |        143 |
| IRC SSL       |       6670 |
| SIP           | 5060..5061 |
| SMTP          |         25 |
| SMTPS         |        465 |
| SSH           |         22 |
| XMPP          | 5222..5223 |
| XMPP (server) |       5269 |
| XMPP (BOSH)   | 5280..5281 |
| Bitmessage    |       8444 |
| Convergence   | 8432..8433 |

* Hints and Tips
** Messaging security
If you're connected to other friends via Friendica then the preferred way to send private messages is via Friendica's built-in messaging system.  This is a lot more convenient than using GPG with ordinary email and yet still provides a similar level of protection from unwarranted interception.
** Moving Domains
If you're moving servers and using a different domain name or path then you can search and replace URLs within files in the following way:

#+BEGIN_SRC: bash
find /var/www/mynewdomain/htdocs -type f -exec sed -i 's@myolddomain@mynewdomain@g' {} \;
#+END_SRC

If you're moving the blog to a new domain then you will need to delete the lock file:

#+BEGIN_SRC: bash
rm /var/www/myblogdomainname.com/htdocs/fp-content/%%setup.lock
#+END_SRC

Then visit your blog and reinstall it.  Your existing content will be unaffected but you will need to delete the welcome post which gets added and also re-select your chosen theme.

** MySql foo
*** Backup all databases
To back up all mysql databases:

#+BEGIN_SRC: bash
mysqldump -u root -p --all-databases --events > /var/backups/databasebackup.sql
#+END_SRC
*** Restoring a particular mysql database
To restore yesterday's friendica backup:

#+BEGIN_SRC: bash
mysql -D friendica -o < /var/backups/friendica_daily.sql
#+END_SRC

To restore yesterday's mediawiki backup:

#+BEGIN_SRC: bash
mysql -D wikidb -o < /var/backups/wikidb_daily.sql
#+END_SRC
*** Removing mysql server

If you manage to screw up sql server completely then it can be fully deleted with:

#+BEGIN_SRC: bash
ps aux | grep mysql
#+END_SRC

and use /kill -9 <pid>/ to kill all mysql processes.

#+BEGIN_SRC: bash
apt-get remove --purge mysql\*
apt-get clean
updatedb
#+END_SRC
** Regenerating SSL certificates
If a security vulnerability arrises which requires you to regenerate your SSL certificates, such as [[http://filippo.io/Heartbleed]["heartbleed"]], then this can be done as follows:

Obtain the latest updates:

#+BEGIN_SRC: bash
apt-get update
apt-get upgrade
#+END_SRC

Run *makecert <domain>* for each of your sites.

Recreate the XMPP certificate:

#+BEGIN_SRC: bash
openssl genrsa -out /etc/ssl/private/xmpp.key 4096
openssl req -new -x509 -key /etc/ssl/private/xmpp.key -out /etc/ssl/certs/xmpp.crt -days 3650
chmod 600 /etc/ssl/private/xmpp.key
chmod 600 /etc/ssl/certs/xmpp.crt
chown prosody:prosody /etc/ssl/private/xmpp.key
chown prosody:prosody /etc/ssl/certs/xmpp.crt
#+END_SRC

And regenerate the IRC server keys:

#+BEGIN_SRC: bash
openssl genrsa -out /etc/ircd-hybrid/key/ircd.key 4096
openssl req -new -x509 -key /etc/ircd-hybrid/key/ircd.key -out /etc/ircd-hybrid/key/ircd.pem -days 3650
chmod 600 /etc/ircd-hybrid/key/ircd.key
chmod 600 /etc/ircd-hybrid/key/ircd.pem
#+END_SRC

As an added precaution you may wish to regenerate your ssh host keys:

#+BEGIN_SRC: bash
rm /etc/ssh/ssh_host_*
dpkg-reconfigure openssh-server
#+END_SRC

Then reboot the server with:

#+BEGIN_SRC: bash
reboot
#+END_SRC

** Example crontab file

This is an example of what your crontab file might look like, with the more frequently run tasks at the top.  For the two most frequent tasks specific minutes within each hour are given and they're arranged to try to minimise the number of things running simultaneously.

#+BEGIN_SRC: bash
# /etc/crontab: system-wide crontab
# Unlike any other crontab you don't have to run the `crontab'
# command to install the new version when you edit this file
# and files in /etc/cron.d. These files also have username fields,
# that none of the other crontabs do.

SHELL=/bin/sh
PATH=/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin

# m            h dom mon dow user command
10,20,30,40,50 * *   *   *   root /usr/bin/timeout 120 /usr/bin/dynamicdns && /usr/bin/spamfilter myusername
15,35,55       * *   *   *   root cd /var/www/mydomainname/htdocs; /usr/bin/timeout 240 /usr/bin/php include/poller.php
17             * *   *   *   root cd / && run-parts --report /etc/cron.hourly
25             6 *   *   *   root test -x /usr/sbin/anacron || ( cd / && run-parts --report /etc/cron.daily )
47             6 *   *   7   root test -x /usr/sbin/anacron || ( cd / && run-parts --report /etc/cron.weekly )
52             6 1   *   *   root test -x /usr/sbin/anacron || ( cd / && run-parts --report /etc/cron.monthly )
#+END_SRC

** Using your own domain
Suppose that you have bought a domain name (rather than using a free subdomain on freedns) and you want to use that instead.

Remove any existing nameservers for your domain (or select "custom" nameservers), then add:

#+BEGIN_SRC: bash
NS1.AFRAID.ORG
NS2.AFRAID.ORG
NS3.AFRAID.ORG
NS4.AFRAID.ORG
#+END_SRC

It might take a few minutes for the above change to take effect.  Within freedns click on "Domains" and add your domains (this might only be available to paid members).  Make sure that they're marked as "private".

Select "Subdomains" from the menu on the left then select the MX entry for your domain and change the destination to *10:mydomainname* rather than *10:mail.mydomainname*.

To route email to one of your freedns domains:

#+BEGIN_SRC: bash
emacs /etc/mailname
#+END_SRC

Add any extra domains which you own, then save and exit.

#+BEGIN_SRC: bash
emacs /etc/exim4/update-exim4.conf.conf
#+END_SRC

Within dc_other_hostnames add your extra domain names, separated by a colon ':' character.

Save and exit, then restart exim.

#+BEGIN_SRC: bash
update-exim4.conf.template -r
update-exim4.conf
service exim4 restart
#+END_SRC

You should now be able to send an email from /postmaster@mynewdomainname/ and it should arrive in your inbox.

** Obtaining an "official" SSL certificate
You can obtain a free "official" (as in recognised by default by web browsers) SSL certificate from [[https://www.startssl.com/][StartSSL]].  You will first need to have bought a domain name, since it's not possible to obtain one for a freedns subdomain, so see [[Using your own domain]] for details of how to do that.  You should also have tested that you can send email to the domain and receive it on the BBB (via Mutt or any other email client).

When creating a SSL certificate it's important that the private key (the private component of the public/private pair in [[https://en.wikipedia.org/wiki/Public-key_cryptography][public key cryptography]]) be generated on the BBB /and remain there/.  Don't generate the private key via the StartSSL certificate wizard because this means that potentially they may retain a copy of it which could then be exfiltrated either via [[https://en.wikipedia.org/wiki/Lavabit][Lavabit]] style methodology, "implants", compromised sysadmins or other "side channel" methods.  So that the private key isn't broadcast on the internet we can instead generate a certificate request, which is really just a request for authorisation of a public key.

Firstly you should have an Apache web site configutaion ready to go.  See [[Setting up a web site]] for details.

Within StartSSL under the validations wizard validate your domain, which means sending an email to it and confirming a code.

Now we can generate the certificate request as follows.

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
openssl genrsa -out /etc/ssl/private/$HOSTNAME.key 2048
chown root:ssl-cert /etc/ssl/private/$HOSTNAME.key
chmod 440 /etc/ssl/private/$HOSTNAME.key
mkdir /etc/ssl/requests
#+END_SRC

Now make a certificate request as follows.  You should copy and paste the whole of this, not just line by line.

#+BEGIN_SRC: bash
openssl req -new -key /etc/ssl/private/$HOSTNAME.key -out /etc/ssl/requests/$HOSTNAME.csr
#+END_SRC

For the email address it's a good idea to use postmaster@mydomainname.

Use a random 20 character password, and keep a note of it.  We'll remove this later.

View the request with:

#+BEGIN_SRC: bash
cat /etc/ssl/requests/$HOSTNAME.csr
#+END_SRC

You can then click on "skip" within the StartSSL certificates wizard and copy and paste the encrypted request into the text entry box.  A confirmation will be emailed back to you normally within a few hours.

Log into your StartSSL account and select *Retrieve Certificate* from the *Tool Box* tab.  Copy the text.

#+BEGIN_SRC: bash
emacs /etc/ssl/certs/$HOSTNAME.crt
#+END_SRC

Paste the public key, then save and exit.  Then on the BBB.

#+BEGIN_SRC: bash
mkdir /etc/ssl/roots
mkdir /etc/ssl/chains
wget "http://www.startssl.com/certs/ca.pem" --output-document="/etc/ssl/roots/startssl-root.ca"
wget "http://www.startssl.com/certs/sub.class1.server.ca.pem" --output-document="/etc/ssl/chains/startssl-sub.class1.server.ca.pem"
wget "http://www.startssl.com/certs/sub.class2.server.ca.pem" --output-document="/etc/ssl/chains/startssl-sub.class2.server.ca.pem"
wget "http://www.startssl.com/certs/sub.class3.server.ca.pem" --output-document="/etc/ssl/chains/startssl-sub.class3.server.ca.pem"
ln -s "/etc/ssl/roots/startssl-root.ca" "/etc/ssl/roots/$HOSTNAME-root.ca"
ln -s "/etc/ssl/chains/startssl-sub.class1.server.ca.pem" "/etc/ssl/chains/$HOSTNAME.ca"
cp "/etc/ssl/certs/$HOSTNAME.crt" "/etc/ssl/certs/$HOSTNAME.crt+chain+root"
test -e "/etc/ssl/chains/$HOSTNAME.ca" && cat "/etc/ssl/chains/$HOSTNAME.ca" >> "/etc/ssl/certs/$HOSTNAME.crt+chain+root"
test -e "/etc/ssl/roots/$HOSTNAME-root.ca" && cat "/etc/ssl/roots/$HOSTNAME-root.ca" >> "/etc/ssl/certs/$HOSTNAME.crt+chain+root"
#+END_SRC

To avoid any possibility of the certificates being accidentally overwritten by self-signed ones at a later date you can create backups.

#+BEGIN_SRC: bash
mkdir /etc/ssl/backups
mkdir /etc/ssl/backups/certs
mkdir /etc/ssl/backups/private
cp /etc/ssl/certs/$HOSTNAME* /etc/ssl/backups/certs/
cp /etc/ssl/private/$HOSTNAME* /etc/ssl/backups/private/
chmod -R 400 /etc/ssl/backups/certs/*
chmod -R 400 /etc/ssl/backups/private/*
#+END_SRC

Remove the certificate password, so if the server is rebooted then it won't wait indefinitely for a non-existant keyboard user to type in a password.

#+BEGIN_SRC: bash
openssl rsa -in /etc/ssl/private/$HOSTNAME.key -out /etc/ssl/private/$HOSTNAME.new.key
cp /etc/ssl/private/$HOSTNAME.new.key /etc/ssl/private/$HOSTNAME.key
shred -zu /etc/ssl/private/$HOSTNAME.new.key
#+END_SRC

Edit your Apache configuration file.

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/$HOSTNAME
#+END_SRC

Add the following to the section which starts with *<VirtualHost *:443>*

#+BEGIN_SRC: bash
	SSLCertificateChainFile /etc/ssl/chains/startssl-sub.class1.server.ca.pem
#+END_SRC

Save and exit, then restart apache.

#+BEGIN_SRC: bash
service apache2 restart
#+END_SRC

Now visit your web site at https://mydomainname.com and you should notice that there is no certificate warning displayed.  You will now be able to install systems which don't allow the use of self-signed certificates, such as [[https://redmatrix.me/&JS=1][Red Matrix]].

* Deprecated

The following items have been deprecated until such time as a successful installation is achieved.

** Install a VoIP server

#+BEGIN_VERSE
/Our core principles, whether in software or sovereignty, have always been about freedom and dignity, for all people, on an equal basis/

-- David Sugar, GNU Telephony
#+END_VERSE

*** The server
Sipwitch is like an introduction service or phone book for SIP VoIP clients.  Once introduced the clients can then talk directly, and this means that sipwitch is very lightweight and can run on low power systems such as the BBB.

Edit your package sources:

#+BEGIN_SRC: bash
emacs /etc/apt/sources.list
#+END_SRC

Append the following line:

#+BEGIN_SRC: bash
deb http://dev.gnutelephony.org/archive/ wheezy/
#+END_SRC

Save and exit.

To load the repository the first time after adding it to the sources.list, since you do not have the verification keys already installed yet. Then do

#+BEGIN_SRC: bash
apt-get install gnutelephony-keyring
#+END_SRC

After that it will be happy to accept it as a signed repository. The verification keys can also be directly fetched with

#+BEGIN_SRC: bash
cd /tmp
wget http://dev.gnutelephony.org/archive/wheezy/public.key
#+END_SRC

and manually added instead with

#+BEGIN_SRC: bash
apt-key add public.key
#+END_SRC

To make sure you have all dependencies, do

#+BEGIN_SRC: bash
apt-get update;apt-get dist-upgrade
#+END_SRC

Before we install anything, let's inspect what is available to us by using

#+BEGIN_SRC: bash
dpkg -l sipwitch
#+END_SRC

To see the main application. The columns will indicate if the package is installed, which version and a description of the package. Then do

#+BEGIN_SRC: bash
dpkg -l sipwitch-*
#+END_SRC

to see available supporting applications and plugins. Again, the columns will indicate if the package is installed, which version and a description of each of these.

To install only the main application, do

#+BEGIN_SRC: bash
apt-get install sipwitch
#+END_SRC

and to install all supporting plugins:

#+BEGIN_SRC: bash
apt-get install sipwitch-plugin-scripting sipwitch-plugin-subscriber sipwitch-plugin-forward sipwitch-plugin-zeroconf
#+END_SRC

Add your user into the sipwitch group

#+BEGIN_SRC: bash
groupadd sipwitch
groupadd sipusers
usermod -aG sipwitch myusername
usermod -aG sipusers myusername
#+END_SRC

Then edit the configuration

#+BEGIN_SRC: bash
emacs /etc/sipwitch.conf
#+END_SRC

Change the *mapped* value from 200 to 20, since we don't want to be serving huge numbers of calls.

Alter the *range* value to 10, since we don't need a large number of extensions.  This will mean that exension numbers 200 to 209 are available.

Do not set the *realm* value, as doing so seems to prevent the server from working.

Save and exit.

Create a digest string for your username:

#+BEGIN_SRC: bash
sipwitch digest myusername
#+END_SRC

Make a note of the resulting string because you're going to use it in the users file you'll now create.

#+BEGIN_SRC: bash
export HOSTNAME=mydomainname.com
touch /etc/sipwitch.d/$HOSTNAME.xml
chmod 600 /etc/sipwitch.d/$HOSTNAME.xml
emacs /etc/sipwitch.d/$HOSTNAME.xml
#+END_SRC

It should look something like the following:

#+BEGIN_SRC: bash
<provision>
    <user id="myusername">
        <digest>yourdigeststring</digest>
		<extension>201</extension>
		<display>Your full name</display>
    </user>
</provision>
#+END_SRC

Save and exit.  Now edit the configuration.

#+BEGIN_SRC: bash
emacs /etc/default/sipwitch
#+END_SRC

Change "desktop" to "server", then save and exit.

Update the IP settings:

#+BEGIN_SRC: bash
iptables -A INPUT -p tcp --dport 5060 -j ACCEPT
iptables -A INPUT -p udp --dport 5060 -j ACCEPT
iptables -A INPUT -p tcp --dport 5061 -j ACCEPT
iptables -A INPUT -p udp --dport 5061 -j ACCEPT
iptables-save
#+END_SRC

Test that it's working:

#+BEGIN_SRC: bash
pkill -9 sipw
sipw -x9 -f
#+END_SRC

Then try to register with the server using a SIP client (such as Jitsi). If everything worked then use CTRL-C to exit.  Then start the service.

#+BEGIN_SRC: bash
service sipwitch start
#+END_SRC

*** Clients
**** Jitsi
Download the latst version from https://jitsi.org/index.php/Main/Download

TODO
**** Twinkle client

The client should have a user profile as following:

The "user name" is the xxx id used in the <user id="xxx"> entry of /etc/sipwitch.conf

The "domain" is the yyy domain in the main config <stack><domain>yyy entry of /etc/sipwitch.conf

The SIP Authentication should have:

realm = realm as set in <registry><realm> of /etc/sipwitch.conf

authentication name = <user id="xx"> entry, same as "User Name" field.

password = value of <secret>zzz in <user> entry of /etc/sipwitch.conf

Under security tab, set "Enable ZRTP/SRTP encryption"
**** Android
TODO

CSipSimple?
** Kune
Kune is a collaboration tool aimed at not just socialising but also getting stuff done within a community.  It's based upon Apache Wave (formerly Google Wave).

#+BEGIN_SRC: bash
apt-get install openjdk-6-jdk openjdk-7-jre mysql-server adduser dbconfig-common libjmagick6-jni
#+END_SRC

Add the Kune repository:

#+BEGIN_SRC: bash
emacs /etc/apt/sources.list
#+END_SRC

Append the following:

#+BEGIN_SRC: bash
deb ftp://ftp.kune.ourproject.org/pub/kune/debian/ stable/
#+END_SRC

Save and exit, then install the Kune package.

#+BEGIN_SRC: bash
gpg --keyserver pgp.mit.edu --recv-keys 9E358A05
gpg --armor --export 9E358A05 | apt-key add -
apt-get update
apt-get install kune
#+END_SRC

You will be asked for the MySql root password and another password to be used with the Kune database.

Allow the system to start automatically at boot.

#+BEGIN_SRC: bash
emacs /etc/default/kune
#+END_SRC

Set /START=yes/, then save and exit.

#+BEGIN_SRC: bash
service kune start
#+END_SRC

Now configure Apache.

#+BEGIN_SRC: bash
a2enmod expires
a2enmod proxy
a2enmod proxy_connect
a2enmod proxy_http
#+END_SRC

Upgrade the database.

#+BEGIN_SRC: bash
mysql -p kune_prod < /usr/share/dbconfig-common/data/kune/upgrade/mysql/0.1.0+b5
mysql -p kune_prod < /usr/share/dbconfig-common/data/kune/upgrade/mysql/0.1.0+b6
mysql -p kune_prod < /usr/share/dbconfig-common/data/kune/upgrade/mysql/0.2.0+b12
mysql -p kune_prod < /usr/share/dbconfig-common/data/kune/upgrade/mysql/0.2.0+b23
mysql -p kune_prod < /usr/share/dbconfig-common/data/kune/upgrade/mysql/0.2.0+b25
#+END_SRC

Edit the Apache configuration.

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/$HOSTNAME
#+END_SRC

<VirtualHost *:80>
    ServerName YOURSERVERNAME

    ProxyRequests Off
    <Proxy *>
        Order deny,allow
        Allow from all
    </Proxy>

    <Files *.cache.*>
        ExpiresActive On
        ExpiresDefault "modification plus 2 years"
    </Files>

    <Files *.nocache.*>
        ExpiresActive Off
    </Files>

    ProxyPass /kune/ http://localhost:8888/
    ProxyPassReverse /kune/ http://localhost:8888/
    <Location /var/www/mydomainname.com/htdocs/kune/>
        Order allow,deny
        Allow from all
    </Location>
</VirtualHost>


Within a browser open https://mydomainname.com:8888

See documentation in /usr/share/doc/kune/INSTALL.gz

** Loomio

#+BEGIN_SRC: bash
apt-get install imagemagick libmagickcore-dev postgresql libmagickwand-dev
#+END_SRC

psql -d postgres
postgres=# create role postgres login createdb;
postgres=# \q


#+BEGIN_SRC: bash
cd /srv
git clone https://github.com/loomio/loomio.git
cd /srv/loomio
bundle install
cp config/database.example.yml config/database.yml
cp .example-env .env
bundle exec rake db:create
bundle exec rake db:schema:load
bundle exec rake db:schema:load RAILS_ENV=test
bundle exec rake db:seed
#+END_SRC

foreman start

Edit the Apache configuration for your mediagoblin site.

#+BEGIN_SRC: bash
emacs /etc/apache2/sites-available/myloomiodomain
#+END_SRC

Delete the existing configuration (in Emacs it's CTRL-x h then CTRL-w) and paste the following, replacing /myloomiodomain/ with your mediagoblin domain name and /myusername@mydomainname.com/ with your email address.

#+BEGIN_SRC: bash
<VirtualHost *:80>
    ServerAdmin myusername@mydomainname.com

    DocumentRoot /srv/myloomiodomain
    ServerName myloomiodomain

    <Directory />
        Options FollowSymLinks
        AllowOverride None
    </Directory>
    <Directory /srv/myloomiodomain/>
        Options Indexes FollowSymLinks MultiViews
        AllowOverride All
        Order allow,deny
        allow from all
    </Directory>

    LogLevel warn

    ProxyVia On

    ProxyRequests off
    ProxyPreserveHost on

    ProxyPass / http://localhost:3000/

    ErrorLog "/var/log/apache2/error.log"
    CustomLog "/var/log/apache2/access.log" combined

    RewriteEngine On
    RewriteOptions Inherit
</VirtualHost>
#+END_SRC

Save and exit.

Now in a browser visit http://myloomiodomain and create a user.

* Related projects

  * [[https://freedomboxfoundation.org/][Freedombox]]
  * [[https://arkos.io/][ArkOS]]